Two-layer optimization decoupling method, device and computer equipment based on proxy function

Abstract

The present application relates to a two-layer optimization decoupling method, device and computer equipment based on a proxy function, wherein the method includes obtaining the current mission flight plan parameters and maintenance task parameters; calling the constructed DP‑ASRS dynamic model; optimizing and solving the DP‑ASRS dynamic model according to the mission flight plan parameters and maintenance task parameters using a two-stage optimization method, decoupling and outputting the current integrated scheduling solution; the integrated scheduling solution is used for integrated scheduling of distributed platform fleets. Among them, the objective function of the DP‑ASRS dynamic model includes: the distributed platform mission comprehensive effectiveness optimization objective function and the single platform effectiveness optimization objective function, and the constraints of the DP‑ASRS dynamic model include distributed platform carrier-based aircraft fleet allocation constraints, flight timing planning constraints, maintenance scheduling constraints and distributed maintenance task platform transfer constraints. The purpose of efficiently realizing the integrated scheduling of distributed platform fleets is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of platform cluster scheduling, and in particular to a two-layer optimization decoupling method, device and computer equipment based on proxy functions. Background Art

[0002] Carrier-based aircraft are the main combat units participating in the combat and training missions of offshore platforms. The dispatching and maintenance capabilities of carrier-based aircraft affect the performance of the platform's combat effectiveness. The dispatching flight plan of carrier-based aircraft must meet the mission requirements of large-scale, high-intensity and continuous combat, and is highly coupled with the maintenance plan. The offshore platform hangar where the maintenance mission is located has the characteristics of small space, complex environment, high degree of coordination of maintenance personnel, limited staffing of maintenance personnel, and complex maintenance process flow, which makes scheduling and planning difficult. At the same time, compared with the scheduling and planning of a single platform, the number of decision-making aircraft groups on distributed platforms is large and the status is complex. In recent years, many achievements have emerged in the phased operation scheduling of carrier-based aircraft, including several operation stages such as deployment and transportation, ammunition support, maintenance and service support, dispatching and departure, landing and recovery, and maintenance and repair.

[0003] The scheduling of carrier-based aircraft operations is based on flight plans. In traditional sortie missions, the commander first formulates a sortie flight plan through experience, and then optimizes and schedules the subsequent phased operation processes in sequence. The formulation of a sortie flight plan can directly affect the optimization effect of subsequent operation scheduling. However, most existing studies are based on a given mission plan for optimization design, and the flight plan for carrier-based aircraft is usually a known condition, and there is no relevant research at present. However, in the process of realizing the present invention, the inventor found that with the increase in mission complexity and the development trend of distributed operations, it is no longer possible to efficiently achieve the integrated scheduling of distributed platform fleets by manual experience alone. Summary of the invention

[0004] Based on this, it is necessary to provide a two-layer optimization decoupling method based on proxy function, a two-layer optimization decoupling device based on proxy function and a computer device to address the above technical problems, which can efficiently realize the integrated scheduling of distributed platform clusters.

[0005] In order to achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0006] On the one hand, an embodiment of the present invention provides a two-layer optimization decoupling method based on a proxy function, comprising the steps of:

[0007] Obtain the current mission flight plan parameters and maintenance mission parameters; the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft on the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single-wave deck service support duration, the platform impact weight, and the performance evaluation index impact weight; the maintenance mission parameters include the number of maintenance skills, the number of platform maintenance personnel, and the maintenance mode;

[0008] The constructed DP-ASRS dynamic model is called; the objective functions of the DP-ASRS dynamic model include: the distributed platform task comprehensive effectiveness optimization objective function and the single platform effectiveness optimization objective function; the constraints of the DP-ASRS dynamic model include the distributed platform carrier-based aircraft fleet allocation constraints, flight timing planning constraints, maintenance scheduling constraints and distributed maintenance task platform transfer constraints;

[0009] According to the mission flight plan parameters and maintenance task parameters, the DP-ASRS dynamic model is optimized and solved using a two-stage optimization method, and the current integrated scheduling scheme is decoupled and output; the integrated scheduling scheme is used for integrated scheduling of distributed platform fleets.

[0010] On the other hand, a two-layer optimization decoupling device based on a proxy function is also provided, comprising:

[0011] The data acquisition module is used to obtain the current mission flight plan parameters and maintenance mission parameters; the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft on the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single-wave deck service support duration, the platform impact weight and the efficiency evaluation index impact weight; the maintenance mission parameters include the number of maintenance skills, the number of platform maintenance personnel and the maintenance mode;

[0012] The model calling module is used to call the constructed DP-ASRS dynamic model; the objective functions of the DP-ASRS dynamic model include: the distributed platform task comprehensive effectiveness optimization objective function and the single platform effectiveness optimization objective function; the constraints of the DP-ASRS dynamic model include the distributed platform carrier-based aircraft fleet allocation constraints, flight timing planning constraints, maintenance scheduling constraints and distributed maintenance task platform transfer constraints;

[0013] The optimization output module is used to optimize and solve the DP-ASRS dynamic model according to the mission flight plan parameters and maintenance task parameters using a two-stage optimization method, and decouple and output the current integrated scheduling plan; the integrated scheduling plan is used for integrated scheduling of distributed platform fleets.

[0014] On the other hand, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned two-layer optimization decoupling method based on the proxy function can be implemented.

[0015] One of the above technical solutions has the following advantages and beneficial effects:

[0016] The above-mentioned two-layer optimization decoupling method, device and computer equipment based on proxy function first obtain the current mission flight plan parameters and maintenance task parameters for the currently required integrated scheduling of the distributed platform fleet, and then call the DP-ASRS dynamic model constructed for the distributed platform, and use the acquired known parameter data to optimize and solve the DP-ASRS dynamic model through a two-stage optimization method, decouple and output the distributed platform task comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function to maximize the optimal scheduling plan that meets the constraints of the model, that is, the current integrated scheduling plan is obtained, and finally the integrated scheduling plan can be used to carry out the integrated scheduling of the distributed platform fleet, thereby achieving the purpose of efficiently realizing the integrated scheduling of the distributed platform fleet, and having strong real-time and applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 This is a schematic diagram of double-shift continuous dispatch;

[0019] Figure 2 A schematic diagram of the platform's flight plan;

[0020] Figure 3 This is a schematic diagram of the carrier-based aircraft maintenance plan;

[0021] Figure 4 A schematic diagram of reactive rescheduling in one embodiment;

[0022] Figure 5 It is a flowchart of a two-layer optimization decoupling method based on a proxy function in one embodiment;

[0023] Figure 6 A flowchart of a single machine maintenance network in an embodiment;

[0024] Figure 7Schematic diagram of setting initial values ​​of accumulated flight time for three groups of cases in one embodiment, (a) is the distribution of initial values ​​of accumulated flight time, (b) is the distribution of available time before the next scheduled maintenance;

[0025] Figure 8 This is a Gantt chart of the dual-layer optimization scheduling of platform 3 in case 2 in an embodiment, (a) is the Gantt chart of the wave dispatch flight sequence, and (b) is the Gantt chart of the maintenance plan;

[0026] Fig. 9 This is a schematic diagram of the allocation result of the platform 3 fleet in an embodiment, (a) is the cumulative flight time allocation result, and (b) is the available time distribution allocation result before the next scheduled maintenance;

[0027] Fig.10 Distributing proxy functions based on a distributed platform cluster in one embodiment The distribution result diagram of ;

[0028] Fig.11 A schematic diagram showing the comparison of comprehensive performance evaluation index results for Case 2 in an embodiment;

[0029] Fig.12 This is a baseline scheduling Gantt chart of platform 2 in case 2 in an embodiment, (a) is a wave dispatch flight sequence Gantt chart, and (b) is a maintenance plan Gantt chart;

[0030] Fig.13 This is a Gantt chart of independent maintenance reactive scheduling of platform 2 in case 2 in an embodiment, (a) is a Gantt chart of wave dispatch flight sequence, and (b) is a Gantt chart of maintenance plan;

[0031] Fig.14 This is a Gantt chart of cross-platform maintenance linkage reactive scheduling for Case 2 Platform 2 in an embodiment, (a) is a Gantt chart of the wave dispatch flight sequence, and (b) is a Gantt chart of the maintenance plan;

[0032] Fig.15 The figure is a schematic diagram of the module structure of a double-layer optimization decoupling device based on a proxy function in one embodiment. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application. The terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The term "and / or" used herein includes any and all combinations of one or more related listed items. In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the ability of ordinary technicians in the field to achieve it. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0034] In response to the above-mentioned technical problem of how to efficiently realize the integrated scheduling of distributed platform fleets, this application takes the flight plan and maintenance scheduling of the wave sortie of the maritime distributed platform fleet as the research object, systematically analyzes the flight timing planning, cyclic sortie operation rules and maintenance constraint characteristics of carrier-based aircraft, and introduces a cross-platform maintenance linkage mechanism based on the distributed maintenance task allocation heuristic rules based on the traditional single-platform independent maintenance form. The distributed platform fleet sortie and repair integrated scheduling (Aircraft fleet Sortie and Repair Integrated Scheduling of Distributed Platforms, DP-ASRS) optimization model is studied and established. Aiming at the indicators such as the comprehensive effectiveness of distributed platform tasks, single platform effectiveness, task wave completion rate, maintenance load dispersion and wave rotation continuity, the scientific allocation of carrier-based aircraft fleets, maintenance timing and maintenance tasks is realized, and the evaluation indicators such as the task wave completion rate and maintenance load dispersion are effectively improved, and the comprehensive effectiveness of combat and training tasks is further improved. Provide a reliable and realistic basis for the integrated planning and scheduling decision-making of flight plans and maintenance tasks for future distributed maritime combat carrier-based aircraft fleets.

[0035] The implementation modes of the present invention will be described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

[0036] Cyclic sortie operation: The sortie and recovery operation style of carrier-based aircraft can be divided into two modes: wave operation and continuous operation. The wave operation mode is to dispatch and recover waves according to the flight cycle (take-off-landing), and there is a preparation cycle between adjacent flight cycles. The flight cycle of the continuous operation mode is continuous, and adjacent flight cycles are staggered and overlapped. The most commonly used form for the continuous operation mode is the single-cycle double-wave cyclic formation, that is, "double-shift cyclic continuous sortie". With one flight cycle as the cycle unit, two formations are selected from the available carrier-based aircraft on the platform, and they are alternately dispatched and recovered in the order of waves, and the cycle is carried out (such as the first wave, the third wave, the fifth wave... The first formation of carrier-based aircraft group is dispatched, the second wave, the fourth wave, the sixth wave... The second formation of carrier-based aircraft group is dispatched), until the set number of continuous rotation waves is reached, the two formations are released, and two formations are selected again to continue the cycle to achieve continuous operation. The whole process has the characteristics of cyclic cycle and rhythmic law. The alternating cycle operation pattern can be flexibly formulated according to the deck operation cycle, the number of continuous rotation waves, the number of carrier-based aircraft in a single wave, the total number of waves, etc. The schematic diagram of double-shift continuous sorties is shown in Figure 1.

[0037] Service support refers to a series of activities such as taxiing to the support parking position after the carrier-based aircraft completes its mission and lands, recharging "oil, liquid, gas and electricity", and adding weapons and ammunition, etc., in order to prepare for the next mission of the carrier-based aircraft and provide necessary guarantees for aviation operation safety and normal operation of the system. Service support preparation before flight needs to be carried out within a deck operation cycle, and the duration of a deck cycle is usually limited to a specified time, such as 1+30 (1 hour and 30 minutes) and 2+00, etc.

[0038] Dispatch flight plan: The flight sequence planning of the fleet is to determine the status of each carrier-based aircraft through the ADMACS system under the task requirements determined by the basic operation cycle, the dispatch operation mode and the single wave flight time. At the beginning of each wave, the required number of carrier-based aircraft for the wave mission is selected from the available fleet to formulate the platform dispatch flight plan as shown in Figure 2. It can be seen that the flight plan diagram not only contains the timing planning such as the number of carrier-based aircraft performing flight missions, but also includes service support, maintenance mode, maintenance start and end time and other information. The maintenance related information reflects the actual maintenance time required, which is related to the process scheduling results of the maintenance task in the hangar, which is the difficulty of flight plan formulation.

[0039] On the one hand, the selection of the fleet for each wave of missions in the flight plan affects the timing of each carrier-based aircraft entering maintenance; on the other hand, entering the maintenance sequence under the condition of an uncertain number of maintenance tasks being carried out in the hangar will affect the maintenance scheduling plan (baseline scheduling plan) currently being executed, requiring a reactive rescheduling of the baseline scheduling plan (as shown in Figure 2). t3). At the same time, the subsequent moment may still cause new carrier-based aircraft to enter the maintenance state due to the flight plan, causing the baseline scheduling plan to be adjusted again. Therefore, affected by the dynamic characteristics, the maintenance time of each aircraft is uncertain and difficult to predict in advance. It is necessary to conduct actual maintenance dynamic scheduling according to the specific situation to obtain the final scheduling plan, and feedback to the ADMACS system for the update of the available fleet set, which in turn affects the selection of the flight plan wave mission fleet. It can be seen that the wave-based flight sequence planning and maintenance scheduling are intertwined and coupled, and integrated scheduling optimization is required.

[0040] Carrier-based aircraft inspection and repair: The offshore platform hangar is responsible for the maintenance and repair of carrier-based aircraft. Maintenance refers to the regular preventive maintenance of carrier-based aircraft, which usually includes periodic inspections of 25 flight hours, 50 flight hours, 100 flight hours, 7 days, 3 months and 6 months. Maintenance is a corrective fault repair work, which refers to repairing faults and damage within a certain limit to restore its original function and availability. Carry out regular maintenance of carrier-based aircraft, troubleshooting and repair of airborne equipment, repair of airframe structure damage and aircraft engine repair to maintain or restore the good technical condition of carrier-based aircraft during their stay on board, maintain a high level of integrity of the fleet, and prepare for wave-based combat training tasks (maintenance and repair are collectively referred to as "inspection" in this article).

[0041] For the maintenance process of carrier-based aircraft: The maintenance tasks of carrier-based aircraft have simple serial relationships among the processes, which are divided into three processes: fault location, fault repair, and re-inspection. For the maintenance tasks of carrier-based aircraft, the processes are networked in sequence, and the immediate predecessor of any process may not be unique. The Activity on Node (AoN) is used to represent the networked relationship between the processes.

[0042] The maintenance scheduling of carrier-based aircraft is to carry out the time sequence planning of the process and the assignment of maintenance personnel or equipment and other resources. The maintenance personnel establishment in the offshore platform hangar is usually set according to the professional category. The maintenance process corresponds to a specific profession, and the personnel are divided into categories according to the profession they master. After the carrier-based aircraft to be repaired enters the hangar maintenance task sequence, the time sequence scheduling plan of the process and personnel that meets the model constraints is planned to generate the carrier-based aircraft maintenance plan diagram as shown in Figure 3. At the same time, the reactive rescheduling mentioned in the previous section needs to be considered (as shown in Figures 2, 3 and 4, respectively). t 3): Based on the baseline scheduling plan before adding the task sequence, the process sets that have been completed, are in progress, and have not yet started scheduling are screened out, the occupied resource status is initialized, the baseline scheduling plan for the completed and in progress scheduling process sets is kept unchanged, and the process sets that have not yet started scheduling (including the newly added maintenance task process) are rescheduled according to the "first come, first served" rule under resource constraints.

[0043] Allocation and maintenance linkage mechanism: The DP-ASRS problem involves two aspects of allocation: one is the allocation of the platform for the fleet; the other is the allocation of cross-platform carrier-based aircraft maintenance tasks during the mission. The allocation of the platform for the fleet refers to the assignment of the carrier-based aircraft to the platform one by one based on the cumulative flight time of the carrier-based aircraft, under the condition that the number of carrier-based aircraft carried by the distributed platform is constrained. The accumulated flight time is the flight status attribute of the carrier-based aircraft. When the flight time of the carrier-based aircraft reaches a certain time, the carrier-based aircraft needs to be transferred to the platform hangar for regular maintenance. The accumulated flight time of the carrier-based aircraft directly affects the timing of the aircraft entering regular maintenance, and indirectly affects the available status during the subsequent mission sortie, that is, it affects the formulation of the wave sortie flight plan. Allocation is the basic premise for formulating the flight plan. The results of the fleet allocation will directly affect the effect of the subsequent wave sortie timing planning stage, and even affect the completion of the combat training mission.

[0044] Random failures or combat damages that occur during the mission of carrier-based aircraft will be transferred to the platform hangar for maintenance after the recovery phase. When the battle is fierce, the concentration of a large number of failed carrier-based aircraft will cause a backlog of maintenance tasks, which will lead to a decrease in the availability of the fleet for a period of time, and the recovery is slow, which will seriously affect the generation of subsequent combat effectiveness. In order to avoid the concentration of maintenance tasks in a single platform hangar, under the concept of distributed maritime operations, a cross-platform maintenance linkage mechanism based on distributed maintenance task allocation heuristic rules is proposed: relying on the powerful heterogeneous network convergence communication of the distributed system to achieve accurate and real-time data sharing of distributed platforms, and realize "ship-aircraft-ship maintenance linkage". Carrier-based aircraft that fail or are damaged during the flight mission receive load status data from the mother ship and other platform hangars before entering the wave recovery phase, and make a comprehensive judgment on the recovery platform according to the distributed maintenance task allocation heuristic rules. When the maintenance task is judged to be transferred to other platforms for maintenance, the faulty carrier-based aircraft will be added to the target platform recovery formation during the recovery phase for cross-platform recovery, thereby achieving the dispersion of platform maintenance load pressure. At the same time, a corresponding number of carrier-based aircraft are selected from the target platform recovery formation and transferred to the faulty aircraft carrier platform for recovery, so that the loss difference of the mother ship can be supplemented.

[0045] See also Figure 5 In one embodiment, the present invention provides a two-layer optimization decoupling method based on a proxy function, comprising the following steps S12 to S16:

[0046] S12, obtaining the current mission flight plan parameters and maintenance mission parameters; the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft on board the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single-wave deck service support duration, the platform impact weight, and the performance evaluation index impact weight; the maintenance mission parameters include the number of maintenance skills, the number of platform maintenance personnel, and the maintenance mode.

[0047] It can be understood that for a given distributed platform in the current actual application scenario (which may include multiple carrier-based aircraft platforms at sea), its mission flight plan parameters and maintenance mission parameters are mission-determined or known data, which can be directly input or directly obtained from the aviation data management and control system ADMACS.

[0048] S14, calling the constructed DP-ASRS dynamic model; the objective functions of the DP-ASRS dynamic model include: distributed platform task comprehensive effectiveness optimization objective function and single platform effectiveness optimization objective function, and the constraints of the DP-ASRS dynamic model include distributed platform carrier-based aircraft fleet allocation constraints, flight timing planning constraints, maintenance scheduling constraints and distributed maintenance task platform transfer constraints.

[0049] It can be understood that the DP-ASRS dynamic model is a model specially constructed for the integrated dispatching and maintenance scheduling requirements of distributed platforms. It is composed of the distributed platform task comprehensive efficiency optimization objective function, the single platform efficiency optimization objective function, and the determined constraints. The DP-ASRS dynamic model can be pre-installed in the data system used for command and dispatch so that it can be quickly called in actual application.

[0050] S16, according to the mission flight plan parameters and maintenance mission parameters, the DP-ASRS dynamic model is optimized and solved using a two-stage optimization method, and the current integrated scheduling plan is decoupled and output; the integrated scheduling plan is used for integrated scheduling of distributed platform fleets.

[0051] It can be understood that according to the mission flight plan parameters and maintenance task parameters, the DP-ASRS dynamic model is optimized and solved using the two-stage optimization method, that is, using the currently acquired mission flight plan parameters and maintenance task parameters, under the constraints of the model, the distributed platform mission comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function are maximized and calculated to obtain the optimal scheduling plan. This optimal scheduling plan is the current integrated scheduling plan, under which the distributed platform mission comprehensive efficiency is maximized and the single platform efficiency is maximized, and the corresponding distributed platform fleet wave dispatch planning and maintenance scheduling plan are given.

[0052] After using the current mission flight plan parameters and maintenance mission parameters to call the DP-ASRS dynamic model to quickly output the current integrated scheduling plan, the commander can use the integrated scheduling plan through the system used to carry out command and dispatch to perform integrated scheduling of the platform fleet for the current distributed platform, so as to maximize the comprehensive effectiveness of combat and training tasks based on distributed platforms.

[0053] The above-mentioned two-layer optimization decoupling method based on proxy function, for the currently required integrated scheduling of distributed platform fleets, first obtains the current mission flight plan parameters and maintenance mission parameters, then calls the constructed DP-ASRS dynamic model, and uses the acquired known parameter data to optimize and solve the DP-ASRS dynamic model through a two-stage optimization method, decouples and outputs the distributed platform task comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function to maximize the optimal scheduling plan that meets the constraints of the model, that is, the current integrated scheduling plan is obtained. Finally, the integrated scheduling plan can be used to carry out the integrated scheduling of distributed platform fleets, thereby achieving the purpose of efficiently realizing the integrated scheduling of distributed platform fleets, with strong real-time and applicability.

[0054] In one embodiment, the construction process of the DP-ASRS dynamic model may be as follows:

[0055] Before the model is established, reasonable assumptions are made based on the objective conditions of the distributed platform: 1) The wave dispatch mission plan is known; 2) The cost of transferring maintenance tasks across platforms is not considered; 3) The time for task allocation and transfer within the hangar and between platforms is ignored; 4) The maintenance process cannot be interrupted after it starts; 5) The platform can perform maintenance and repair on all types of carrier-based aircraft.

[0056] Regarding the definition of model parameters and variables, specifically, the parameter definition of the distributed platform fleet wave dispatch and periodic maintenance integrated static scheduling model is as follows:

[0057] : The set of carrier-based aircraft to be assigned, , is the number of carrier-based aircraft to be allocated. : The pth platform carries a collection of carrier-based aircraft, , is the number of carrier-based aircraft carried by the pth platform. P: the number of carrier-based aircraft carrying platforms, , The number of carrier-based aircraft platforms. : The pth platform mission dispatch wave set, , Indicates the number of mission waves dispatched for the p-th platform. n : No. p Platform iThe number of times the carrier-based aircraft are repaired is a collection. , Indicates p Platform i Number of times carrier-based aircraft were repaired. : No. p Platform i The carrier-based aircraft is executing n The length of time that the device will be available before the next scheduled maintenance. : No. p Platform i The carrier-based aircraft is executing n The theoretical maintenance duration of a scheduled maintenance. : Maintenance timing offset duration, allowing maintenance flight hours before and after Flexibly enter maintenance status within the range. : The duration of the flight cycle of the pth platform. : Duration of the operation cycle of the pth platform deck. : Deck service support duration of a single wave on the pth platform. : The number of continuous rotation waves of the p-th platform cycle operation. : The number of sub-wave groups dispatched in the p-th platform cycle. : The number of carrier-based aircraft required for a single-wave flight mission of the p-th platform. : No. p The carrier-based aircraft on each platform are gathering for maintenance. : No. p Platforms t The carrier-based aircraft can be assembled at any time. , For the p Platforms t The number of carrier-based aircraft available at any given time. : No. p Platforms t Always ready to gather carrier-based aircraft for maintenance. : No. p Platforms t Always waiting for the repair of carrier-based aircraft to gather, , For the p Platforms t The number of carrier-based aircraft awaiting repair at any given time. : No. p Platforms t The number of carrier-based aircraft waiting for maintenance after maintenance tasks are assigned at any time. : No. p Platforms t Always ready to gather carrier-based aircraft for maintenance. : No. p Platform i The carrier-based aircraft is executingn The maintenance process set, . : The total maintenance process set of the fleet, . : No. p The platform cluster is at t A collection of all maintenance operations that are in execution state. : No. p Platform i Carrier-based aircraft at time t A collection of maintenance operations that are in execution state. : No. p Platform i Carrier-based aircraft n The maximum completion time required to complete the maintenance process. : No. p Platform i Carrier-based aircraft n Maintenance No. j Road maintenance process ( i , j )’s immediate predecessor process set. : No. p Platform i Carrier-based aircraft j Road maintenance process ( i , j ) of maintenance work hours. BM : A sufficiently large positive number. K : Overhaul skill set, Kc= {1,2,…,| K |}. : No. p The platform has k A collection of maintenance support personnel with similar maintenance skills. : No. p Platform i Carrier-based aircraft n Maintenance No. j Road maintenance process ( i , j ) k ( k∈Kc ) The demand for such maintenance skills.

[0058] The static decision variables are defined as follows:

[0059] : No. p Platform i Carrier-based aircraft n Maintenance No. j Road maintenance process ( i , j)’s protection starts at that time.

[0060] : No. p Platform i Carrier-based aircraft n Maintenance No. j Road maintenance process ( i , j ) when the guarantee ends.

[0061] .

[0062] .

[0063] .

[0064] .

[0065] .

[0066] In one embodiment, the distributed platform task comprehensive performance optimization objective function is: :

[0067] (1)

[0068] in, express p Platform performance evaluation value, Indicates the influence weight of each platform. Commanders can set the weight based on the mission and platform. P Indicates the number of distributed platforms.

[0069] It is understandable that in the maritime distributed combat training mission, multiple platforms jointly form a combat system, and each platform cooperates with each other, each with its own division of labor, to complete the mission together. With the performance of each single platform as the support, the comprehensive performance optimization goal of the integrated scheduling of the distributed platform fleet is constructed. F As shown in the above formula (1).

[0070] In one embodiment, the single platform performance optimization objective function is: :

[0071] (2)

[0072] in, express p Platform performance evaluation value, , and Indicates the impact weight of the three levels of performance indicators ，P Indicates the number of distributed platforms, represents the task wave completion rate, Indicates the maintenance load distribution, Indicates the continuity of wave rotation.

[0073] It can be understood that the single-platform performance optimization target is defined as shown in the above formula (2), which is reflected in three aspects: task wave completion rate, maintenance load dispersion, and wave rotation continuity.

[0074] Furthermore, the task wave completion rate is:

[0075] (3)

[0076] in, is a task-complete wave variable if and only if hour, Indicates w The wave is a task-completed wave; otherwise Indicates w The wave is a vacant wave. Indicates p Each platform carries a collection of carrier-based aircraft. Indicates the number of carrier-based aircraft required for a single wave flight mission of the platform, Indicates p A set of platform mission dispatch waves, Indicates p The number of mission waves for each platform.

[0077] Specifically, as the flight mission is executed, the available carrier-based aircraft on the platform will successively reach the scheduled maintenance time or fail, resulting in the number of available carrier-based aircraft failing to meet the number requirements of a single wave of carrier-based aircraft. The wave at this time is called a shortfall wave. The proportion of the number of complete waves relative to it has a significant impact on the completion of combat and training missions. The evaluation index of the mission wave completion rate is defined as shown in the above formula (3). The greater the mission wave completion rate, the better the optimization effect of the dispatch flight plan.

[0078] Furthermore, the load distribution is:

[0079] (4)

[0080] in, is the standard deviation calculation function, Indicates p The carrier-based aircraft on each platform are gathering for maintenance.

[0081] Specifically, by reasonably assigning the task sequence of each platform's carrier-based aircraft, dispersing the time for each carrier-based aircraft to enter maintenance, and forming a staggered use interval, we can avoid the concentration of the platform hangar's maintenance load and the reduction of the number of available carrier-based aircraft, which will affect the normal execution of combat and training tasks. The maintenance load dispersion index of the defined sortie flight plan is used to calculate the distribution standard deviation of the time when the platform's carrier-based aircraft enters maintenance, as shown in the above formula (4). The larger the standard deviation, the p The greater the discreteness of the timing of the maintenance of carrier-based aircraft on each platform, the better the optimization effect of the flight plan.

[0082] Furthermore, the wave rotation continuity is:

[0083] (5)

[0084] in, is a continuous rotating wave variable if and only if , When established, Indicates i Ship-borne aircraft Complete continuous rotation within the wave interval, otherwise Indicated in The continuous rotation is not completed within the wave interval; Indicates p Each platform carries a collection of carrier-based aircraft. Indicates the number of carrier-based aircraft required for a single wave flight mission of the platform, Indicates p The platform task dispatch wave set is represented by: Indicates p The number of mission waves per platform, P Indicates the number of distributed platforms, Indicates p The number of continuous rotation waves of the platform cycle operation, Indicates p The number of wave groups dispatched by each platform in a cycle; .

[0085] Specifically, in the cyclic continuous sortie operation mode, the carrier-based aircraft group carried by the platform is divided into several groups, and is dispatched and recovered alternately in wave order. The reason is: on the one hand, considering the continuity of the pilot's dispatch mission, it is required to fly the same aircraft for several waves in succession to maintain the driving habit; on the other hand, it avoids the storage and deck transportation costs involved in the frequent replacement of dispatched aircraft, which is in line with the actual dispatch and recovery operation practices. Maintaining the wave rotation plan is the basis for maintaining the efficiency of support and is of great significance to the combat effectiveness of the aircraft group. The evaluation index of the wave rotation continuity of the cyclic operation is defined as shown in the above formula (5). The greater the continuity of the wave rotation of the cyclic operation, the better the optimization effect of the dispatch flight plan.

[0086] In the formulation of the integrated dispatching plan for fleet wave dispatch and maintenance, the above three levels of performance indicators have certain priorities and are defined. , and To determine the weights of the performance evaluation indicators, commanders can set numerical values ​​based on the importance of the indicators.

[0087] Regarding the constraints of the above model, in one embodiment, the distributed platform carrier-based aircraft fleet allocation constraints include that the number of carrier-based aircraft allocated to the platform matches the number of carrier-based aircraft required for the platform mission and each carrier-based aircraft can only be allocated to one platform in the same allocation.

[0088] Specifically, the allocation constraints of the distributed platform carrier-based aircraft fleet can be expressed as follows:

[0089] (6)

[0090] (7)

[0091] Formula (6) indicates that the number of carrier-based aircraft assigned to the platform matches the number of carrier-based aircraft required for the platform mission; Formula (7) indicates that each carrier-based aircraft can only be assigned to one platform in the same allocation.

[0092] In one embodiment, the flight sequence planning constraints include selection constraints of carrier aircraft wave rotation into adjacent wave formations, mutual exclusion constraints between carrier aircraft mission execution and maintenance / residence, constraints on the number of carrier aircraft dispatched for a single wave flight mission, constraints on the timing of carrier aircraft performing regular maintenance, and constraints on the association of flight sequence planning with maintenance scheduling.

[0093] Specifically, the flight timing planning constraints can be expressed as follows:

[0094] (8)

[0095] (9)

[0096] (10)

[0097] (11)

[0098] (12)

[0099] Formula (8) represents the selection constraint of the wave rotation of carrier-based aircraft, the adjacent wave formation and the wave rotation. The formation and wave rotation requires that any carrier-based aircraft adjacent to the wave selection is mutually exclusive, and the mutual exclusion range is related to the number of wave formations in the cyclic sortie: Related, such as It means the first wave and the second and third waves Mutually exclusive (the values ​​cannot be 1 at the same time). Formula (9) represents the mutually exclusive constraint between the execution of missions and maintenance / residence of carrier-based aircraft. Formula (10) represents the constraint on the number of carrier-based aircraft dispatched for a single wave flight mission, where Indicates p Platform No. ( w -1) The number of carrier-based aircraft occupied by wave missions, Indicates p Platform w The number of carrier-based aircraft that perform regular maintenance in waves, Indicates p Platform w The number of available carrier-based aircraft in a wave. Formula (11) represents the timing constraint for periodic maintenance of carrier-based aircraft. p Platform i The available time for carrier-based aircraft to perform missions is reduced to the time when the n Around the time of scheduled maintenance flight hours Periodic maintenance can be performed within the interval. Formula (12) represents the associated constraints between flight sequence planning and maintenance scheduling. p Platform i The decision variables of the maintenance / staying of carrier-based aircraft and the execution of n The span of the waves of regular maintenance is consistent.

[0100] In one embodiment, the maintenance scheduling constraints include: each maintenance process is carried out in sequence according to the established maintenance process, and the subsequent maintenance task must start after the completion of its immediate predecessor process; each maintenance process has guaranteed start time and guaranteed end time constraints; the skill demand of each maintenance process matches the number of maintenance personnel assigned to each maintenance process and who select this skill; any maintenance personnel can use at most one skill when performing maintenance work on any maintenance process; and priority constraints for maintenance personnel conflicts.

[0101] The distributed maintenance task platform transfer constraints include: the number of repairs to be transferred is less than or equal to the total number of carrier-based aircraft to be repaired; the total number of global transfers is equal to the total number of supplements.

[0102] Specifically, the above constraints can be expressed as follows:

[0103] (13)

[0104] (14)

[0105] (15)

[0106] (16)

[0107] (17)

[0108] Formula (13) indicates that each maintenance process is carried out in sequence according to the established maintenance process, and the subsequent maintenance task must start after the completion of its immediate predecessor process. Formula (14) indicates the guaranteed start time and guaranteed end time constraints for each maintenance process. Formula (15) indicates that the demand for skills in each maintenance process matches the number of maintenance personnel assigned to the process and who choose this skill. Formula (16) indicates that any maintenance personnel can use at most one skill when performing maintenance work on any maintenance process. Formula (17) indicates the priority constraint for maintenance personnel conflicts. When different processes have a demand for the same maintenance personnel, the order of maintenance must be distinguished according to priority.

[0109] The static model of integrated dispatching and maintenance of distributed platform fleet constructed above is shown in formula (18):

[0110] (18)

[0111] When considering the dynamic maintenance of random failures or combat damage, a dynamic scheduling model integrating the wave dispatch and maintenance of distributed platform fleets is constructed, and the decision variables are introduced and defined as follows:

[0112] ;

[0113] .

[0114] The distributed maintenance task platform transfer constraints are introduced as follows:

[0115] (19)

[0116] (20)

[0117] Formula (19) indicates that the number of aircraft to be transferred for repair is less than or equal to the total number of aircraft to be repaired; Formula (20) indicates that the total number of global transfers is equal to the total number of supplements.

[0118] The DP-ASRS dynamic model constructed in combination with the above is shown in formula (21):

[0119] (twenty one)

[0120] In one embodiment, the process of optimizing and solving the DP-ASRS dynamic model using the two-stage optimization method in step S16 includes the following processing steps:

[0121] A. Taking the constructed distributed platform cluster allocation proxy function as the objective function, the evolutionary algorithm is used to solve the distributed platform cluster allocation result optimized in the first stage;

[0122] B. Use the distributed platform cluster allocation results to perform the second phase optimization operation; the second phase optimization operation includes:

[0123] B1. Input the distributed platform cluster allocation result and initialize the population;

[0124] B2. Decode each individual in the population until all task waves are completed;

[0125] B3. Calculate the fitness function value of the individual decoding generation scheme; wherein the fitness function is the single platform performance optimization target;

[0126] B4. Execute population genetic evolution operations;

[0127] B5. Repeat the above steps B2 to B4 until the iteration end condition is met;

[0128] B6. Output the flight plan and performance optimization results of each platform. The population genetic evolution operation in the above step B4 includes:

[0129] The population performs a binary tournament selection operation, selecting individuals with better fitness as parent individuals, generating a parent population of half the size of the population;

[0130] Perform a two-point crossover operation between parent populations to generate offspring populations;

[0131] The offspring population performs random key mutation operations.

[0132] Through the above population genetic evolution operation, high efficiency and good accuracy are achieved.

[0133] It can be understood that in this field, the two-layer optimization method can obtain a better solution for the problem, but the solution efficiency is not high because its framework is an optimization structure of "light on top and heavy on the bottom". The complexity of the upper-layer fleet allocation module is much smaller than that of the lower-layer flight and maintenance timing planning module, and the evaluation of each feasible solution in the optimization process of the upper-layer module needs to call the lower-layer optimization framework, presenting a "nested" structure as a whole. With the increase of the fleet scheduling scale and the number of iterations, the amount of calculation will grow explosively, and the calculation cost is huge.

[0134] In order to solve the problems of low solution efficiency, high cost and poor timeliness of the two-layer optimization method, the above embodiment can be solved by a two-stage optimization method based on a proxy function. In this embodiment, a two-layer optimization decoupling mechanism based on the allocation proxy function is specifically proposed, which transforms the two-layer iteration into a two-stage optimization method, and strives to reduce the complexity of the solution. The second stage corresponds to the lower layer of the two-layer optimization method. The first stage keeps the objective function of the distributed platform cluster allocation module unchanged, and transforms the upper-layer "random allocation-effect feedback-adjustment allocation" optimization mode into a method of using rule function proxy allocation, as shown in the processing steps of the above two stages.

[0135] Furthermore, the constructed distributed platform cluster allocation agent function is as follows:

[0136] ;

[0137] in, , and represents the influence weight of the three rules, Indicates p Each platform carries a collection of carrier-based aircraft. P Indicates the number of distributed platforms, Indicates p Platform i The theoretical maintenance time for a carrier-based aircraft to perform scheduled maintenance is Indicates p Platform i The theoretical maintenance time for a carrier-based aircraft to perform scheduled maintenance is express p The ratio of the average mission sorties of platform carrier-based aircraft to the average mission sorties of distributed platform carrier-based aircraft. express p The ratio of the number of platform maintenance resources configured to the total number of distributed platform maintenance resources.

[0138] It can be understood that the distributed platform cluster allocation agent function The first item is the matching rule between the distributed platform task intensity and the theoretical available time before cluster maintenance, defined as pThe ratio of the average mission sorties of platform carrier-based aircraft to the average mission sorties of distributed platform carrier-based aircraft as follows:

[0139] ;

[0140] It can represent the intensity of tasks undertaken by a single platform. Definition p The ratio of the total available time of the platform cluster before the next scheduled maintenance to the total available time of the distributed platform cluster before the next scheduled maintenance as follows:

[0141] ;

[0142] It can indicate the ability of the fleet on board the platform to continuously perform tasks. In order to meet the high-intensity dispatch missions, the fleet on board the platform should avoid frequent maintenance to ensure high availability. Therefore, the total available time of the platform before the next scheduled maintenance should not be too small. The first function form can satisfy the distributed platform and Match.

[0143] Distributed platform cluster allocation agent function The second item is the matching rule between the distributed platform resource configuration and the theoretical maintenance duration of the cluster, which is defined as p The ratio of the number of platform maintenance resources (maintenance personnel in this article) to the total number of distributed platform maintenance resources as follows:

[0144] ;

[0145] It can represent the maintenance resource allocation occupancy rate of a single platform, and indirectly represent the platform's ability to resist the pressure of maintenance task load; it defines the ratio of the sum of the maximum theoretical maintenance time of a single platform cluster to the sum of the maximum theoretical maintenance time of the total distributed platform cluster. as follows:

[0146] ;

[0147] It can represent the theoretical maintenance pressure of the fleet. The fleet with high theoretical maintenance pressure should be allocated to the platform with high maintenance resource allocation occupancy rate to meet the principle of load balancing of maintenance tasks. The second function form can satisfy the distributed platform and Match.

[0148] Distributed platform cluster allocation agent function The third item is the time control rule for the interval between echelons of the distributed platform fleet. The echelon control of the fleet means that each carrier-based aircraft is sorted according to the available time before the next scheduled maintenance, and a certain interval is maintained between them to present an echelon arrangement. Its purpose is to provide a reference for the carrier-based aircraft to formulate a flight plan, so that each aircraft can disperse the timing of entering maintenance while ensuring the completion of the mission, so as to avoid the concentration of maintenance tasks and affect the normal execution of flight combat training tasks. Fleet allocation also needs to consider fleet echelon control. Indicates allocation to p The part of the sum of the remaining usage time of the front machine and the theoretical maintenance time after the platform fleet is sorted that exceeds the remaining usage time of the rear machine is the theoretical interval in which the maintenance tasks of the two machines can overlap.

[0149] Defining a non-negative function as follows:

[0150] ;

[0151] Combination and , represents the theoretical interval sum of the overlapping maintenance tasks of the distributed platform, It represents the total theoretical maintenance time of the distributed platform cluster. The dimensionless ratio is obtained by dividing the two. The smaller the ratio, the better the control effect of the theoretical interval time of the cluster echelon in the cluster allocation stage. , and , the commander can set the value according to the importance of the rule. The distributed platform allocation for the cluster is performed as the objective function, and a feasible allocation plan can be solved with the help of the above evolutionary algorithm.

[0152] Through the above two-stage optimization method of this embodiment, the current integrated scheduling solution can be accurately and efficiently solved and output, which can greatly reduce the consumption of computing resources.

[0153] In one embodiment, the above step B2 may specifically include the following operations:

[0154] B2.1. Determine whether the available fleet needs maintenance. If so, proceed to B2.2; otherwise, proceed to B2.3.

[0155] B2.2. Execute the first operation; the first operation includes inputting the maintenance mode and maintenance start time, and judging whether to transfer the maintenance task platform according to the distributed maintenance task allocation heuristic rules. If so, execute B2.2.1: execute the maintenance task platform transfer operation, otherwise execute B2.2.2: judge whether to perform reactive rescheduling. Among them, if reactive rescheduling is performed, execute B2.2.3: input the maintenance baseline scheduling result; otherwise execute B2.2.4: initialize the maintenance task priority code according to the "first come, first served" rule; execute B2.2.5: execute the maintenance task decoding operation, and allocate maintenance personnel based on constraints and load balancing principles; execute B2.2.6: save the maintenance baseline scheduling result, and output the maintenance completion time of each carrier-based aircraft. .

[0156] B2.3. Execute the second operation; the second operation includes determining whether the single-wave flight mission is completed, determining whether the maintenance mission is completed, and updating the available fleet;

[0157] B2.4. Selection of carrier-based aircraft for single-wave sorties in compliance with flight sequence planning constraints;

[0158] Repeat steps B2.1 to B2.4 above until all task waves are completed.

[0159] In one embodiment, the heuristic rule for distributed maintenance task allocation requires that after allocation, the number of maintenance task transfers between distributed platforms is minimized when the maintenance task load is balanced between distributed platforms. The defined heuristic rule for distributed maintenance task allocation is:

[0160] ;

[0161] in, is the absolute value function, express p platform t Changes in the number of carrier-based aircraft to be repaired before and after the allocation of maintenance tasks at each moment, express p platform t The maintenance resource occupancy rate at any given moment is defined as:

[0162] ;

[0163] Understandably, The distributed platform maintenance task load balancing can be achieved after task allocation. At the same time, in practical applications, relevant supplementary rules need to be added on the basis of the aforementioned distributed maintenance task allocation heuristic rules: 1. Regular maintenance tasks cannot be transferred; 2. The number selection rules for fault maintenance task transfer and available carrier-based aircraft supplementation are set to prioritize the smaller carrier-based aircraft number by default. 3. The allocation heuristic rules have low complexity and can be optimized by exhaustive search.

[0164] In one embodiment, in order to more intuitively and comprehensively illustrate the above-mentioned two-layer optimization decoupling method based on proxy function, the following is a simulation experiment example of the two-layer optimization decoupling method based on proxy function. It should be noted that the examples given in this specification are only illustrative and are not the only limitation of the specific implementation cases of the present invention. Those skilled in the art can, under the illustration of the implementation cases provided by the present invention, similarly adopt the above-mentioned two-layer optimization decoupling method based on proxy function to realize the integrated scheduling of the wave dispatch and maintenance of the fleet of different distributed platforms.

[0165] Experimental case design: A simulation experiment case was constructed based on three large surface ship platforms distributed at sea performing a high-intensity continuous combat training mission. The simulation was based on MATLAB 2020a software programming and a personal computer (Windows 7 64-bit operating system, Intel(R) Xeon(R) Gold 5122 CPU @ 3.60GHz, 32G memory) was used for calculation. Each algorithm was run independently 20 times and the data was recorded.

[0166] Mission flight plan parameters are set as follows: Mission wave number , the number of distributed platforms , the number of carrier-based aircraft on board the platform , the number of continuous rotation waves of cyclic operation , the number of wave groups for cyclic dispatch , the number of carrier-based aircraft required for a single wave , flight cycle duration , Deck operation cycle time , single wave deck service support time , platform influence weight , the weight of performance evaluation index .

[0167] The maintenance task parameters are set as follows: Maintenance skill quantity K ={1,2,3,4}, set four types of maintenance professional skills: ad hoc, avionics, ordnance and machinery, and the number of maintenance personnel on the three platforms The maintenance modes M1-M6 correspond to 6 maintenance modes: 25h maintenance, 50h maintenance, 100h maintenance, mechanical failure, avionics system failure, and special system failure. The maintenance network process is as follows: Figure 6 As shown, Figure 6 Processes 1 and 20 are virtual start and end processes, the purpose of which is to integrate and connect all processes. The operation time of the carrier-based aircraft maintenance process is shown in Table 1.

[0168] Table 1

[0169]

[0170] The simulation experiment sets three groups of cases with different initial values ​​of the accumulated flight time of the fleet as shown in Figure 7. In the figure, “+” represents the median, " represents the average value, (a) is the distribution of the initial value of the cumulative flight time, and (b) is the distribution of the available time before the next scheduled maintenance. The distribution of the available time before the next scheduled maintenance in Case 3, Case 1 and Case 2 is successively concentrated to the small value, indicating that more and earlier carrier-based aircraft will enter scheduled maintenance in the subsequent cases, and the maintenance task load pressure of the platform hangar increases in sequence. Case 2 can verify the adaptive solution performance of the proposed method for large-scale reactive rescheduling of fleet maintenance tasks and integrated scheduling of fleet wave sorties and maintenance under the condition of highly reduced fleet availability, which is the focus of this experimental case study.

[0171] Comparative test of optimization methods:

[0172] Without considering the random failure of carrier-based aircraft and the cross-platform maintenance linkage mechanism, a two-layer iterative optimization method is used within the static model range of formula (18) to obtain the baseline scheduling scheme for the integrated wave dispatch and scheduled maintenance of distributed carrier-based aircraft groups. Taking Case 2 Platform 3 as an example, the scheduling results are as follows: Figure 8 As shown in the figure, (a) is the Gantt chart of the flight sequence of the wave sortie, and (b) is the Gantt chart of the maintenance plan. Figure 8 The Gantt chart of the optimization results shows that the optimization solution has good continuity in the distribution of maintenance load and wave rotation. During the maintenance plan scheduling process, there were 4 reactive reschedulings, which were at the start of the 4th, 6th, 7th and 9th maintenance. After inspection, the scheduling scheme shown in the Gantt chart meets the constraints of the model, verifying the correctness of the model and the effectiveness of the optimization method.

[0173] The results of cluster allocation on platform 3 are as follows: Fig. 9 As shown, (a) is the cumulative flight time distribution result, and (b) is the available time distribution result before the next scheduled maintenance. Fig. 9 (a) Comparison Figure 7From (a), we can see that the flight time distribution results of platform 3 in the three groups of cases are roughly the same as the overall distribution of the initial values, and there is no obvious regularity between the actual value and distribution of the cumulative flight time. Fig. 9 (b) Comparison Figure 7 (b) shows obvious regular characteristics: the median and average values ​​of the available time before the next scheduled maintenance of the fleet assigned to platform 3 in the three cases are lower than the initial values, and case 3 is more significant. f Among the three groups of cases, case 3 is the largest and case 2 is the smallest. Based on the above results, the optimization of the allocation of the fleet's carrying platforms is not essentially the actual value of the fleet's flight time, but the available time before the next scheduled maintenance of each carrier-based aircraft in the fleet. . The larger the value, the better the WR, MR and CR. F and f The better the performance. Platform 3 has low pressure on flight missions, sufficient maintenance personnel, and no significant demand for available time before scheduled maintenance, which makes the allocation result value lower than the overall level. At the same time, the distribution of available time before the next scheduled maintenance is roughly uniform, and the distribution is hierarchical, which is in line with the expected law. The above results further verify the above distributed platform fleet allocation agent function rationality.

[0174] Table 2

[0175]

[0176] Use distributed platform cluster-based proxy functions The two-stage optimization method is used to solve the problem. Taking Case 2 as an example, we get Fig.10 From the allocation results shown in the above analysis, it can be seen that the allocation of fleet-carrying platforms is actually the allocation of fleets based on the available time before the next scheduled maintenance. The allocation results are brought into the lower-level algorithm for integrated scheduling of distributed fleet wave dispatch and maintenance.

[0177] Taking Case 2 as an example, a comparative experiment between the two-layer iterative optimization method and the two-stage optimization method was conducted. F The comparison results with the optimal solution of the algorithm calculation time are shown in Table 3. f The comparison results with the optimal solutions of the three-level sub-goals WR, MR and CR are as follows: Fig.11 As shown in the figure, comprehensive analysis shows that the two methods have good performance in optimizing indicators. F , f The scheduling results of the optimal solutions of WR, MR and CR are highly consistent, and the optimization goal is F and fThe difference between the optimal values ​​obtained by the two methods is about 1%-2%, while the two-stage optimization method greatly shortens the calculation time. It can be concluded that the comparison results verify the distributed platform cluster allocation agent function. The effectiveness of the two-stage optimization method is demonstrated by using a decoupling mechanism for the double-layer iterative optimization framework. On the basis of ensuring the optimization performance as much as possible, the computing cost is greatly reduced, and the scheduling efficiency is more effectively improved, providing methodological support for the rapid formulation of integrated scheduling plans for the wave dispatch and maintenance of offshore distributed platform fleets.

[0178] Table 3

[0179]

[0180] Cross-platform maintenance linkage mechanism comparison experiment:

[0181] Taking Case 2 as an example, considering the random failure of carrier-based aircraft and the cross-platform maintenance linkage mechanism, firstly, based on the static model of formula (18), a two-stage optimization method is used to solve the integrated baseline scheduling scheme of flight timing planning and scheduled maintenance of distributed carrier-based aircraft groups. Taking platform 2 as an example, the baseline scheduling results are as follows: Fig.12 As shown, (a) is the Gantt chart of the wave flight sequence, and (b) is the Gantt chart of the maintenance plan. Secondly, on the basis of the baseline scheduling plan, random failures or damage based on probability triggering are added to simulate the maritime engagement state of the distributed platform. Dynamic uncertainty disturbances such as failures or damage will affect the execution of the subsequent baseline flight sequence planning. The plan needs to be adjusted reactively after the disturbance. The flight sequence planning after the disturbance adjustment will change locally from the baseline scheduling.

[0182] In order to verify the superiority of the cross-platform maintenance linkage mechanism under the condition of dynamic uncertainty maintenance tasks, a comparison experiment between the cross-platform maintenance linkage mechanism and the independent maintenance of a single platform was conducted. The probability of random failure of the distributed platform carrier-based aircraft is set to 0.1, and the probability of battle damage failure of the platform 1 carrier-based aircraft is set to 0.3. The flight plan and the integrated scheduling scheme of independent maintenance of Case 2 Platform 2 under the condition of independent maintenance of a single platform are calculated. The flight sequence planning results are shown in Figure 2. Fig.13 As shown in (a) in Fig.10 The baseline flight sequence was planned at the beginning of the 11th overhaul. The 7th and 1st carrier-based aircraft had avionics system failures and mechanical failures respectively. The subsequent sortie plan algorithm affected by the failures selected the 28th carrier-based aircraft for a small range of supplementary adjustments to meet the mission requirements. At the same time, Fig.12 (b) The impact of the maintenance plan Gantt chart is adjusted reactively at the same time, and the adjustment result is shown in (b) of Figure 13.

[0183] The DP-ASRS dynamic model based on formula (21) uses a two-stage optimization method to calculate the reactive scheduling adjustment of the cross-platform maintenance linkage mechanism for the baseline scheduling scheme. The maintenance task transfer results are shown in Table 4. Taking the first column of data as an example, it means that at the 24.5h of the mission time, the 26th carrier-based aircraft of platform 1 has an M5 avionics system failure, and the maintenance task is transferred to platform 2. At the same time, the 25th carrier-based aircraft of platform 2 is supplemented to platform 1 to maintain the number of carrier-based aircraft on the platform. Taking platform 2 as an example, the scheduling Gantt chart results of the wave sortie flight sequence are shown in Fig.14 As shown in (a), the dashed box in the figure is marked as the carrier-based aircraft transferred into platform 2, and the solid box is marked as the carrier-based aircraft transferred out of platform 2. It can be seen from the figure that the 26th, 1st (originally carried as platform 3), and 2nd carrier-based aircraft of platform 1 are transferred to platform 2 as maintenance tasks. At the same time, the 25th, 3rd, and 5th carrier-based aircraft of platform 2 are transferred out as supplements of platform 1. The 23rd carrier-based aircraft on platform 2 had a special system failure and was not judged as a maintenance task transferred to other platforms, which is consistent with the data in Table 4. The results of the maintenance plan Gantt chart are shown in the figure. Fig.14 As shown in (b) of Figure 1, combined with the analysis, it can be seen that the complexity of scheduling scale increases greatly after adding random failures or battle damage. During the maintenance plan scheduling process, 13 reactive reschedulings were experienced, which were located at the start of the 2nd, 3rd, 4th, 5th, 7th, 8th, 9th, 11th, 12th, 13th, 14th, 15th and 16th maintenance. At the start of the 9th, 15th and 16th maintenance, maintenance tasks from other platforms were carried out. After inspection, the scheduling scheme shown in the Gantt chart meets the constraints of the model, verifying the correctness of the cross-platform maintenance linkage mechanism in the model.

[0184] Table 4

[0185]

[0186] The complete calculation results of the cross-platform maintenance linkage mechanism and the single-platform independent maintenance scheduling are shown in Table 5. Based on the comparative analysis of the calculation results, the following conclusions are drawn: 1. The cross-platform maintenance linkage mechanism transfers the maintenance task of faulty carrier-based aircraft to the platform with a small maintenance load based on the distributed maintenance task allocation heuristic rule, thereby achieving the dispersion of the maintenance load, which is manifested as an advantage in the MR indicator. 2. For platform 1 with a high probability of failure, under the conditions of Case 1 and Case 2 with high overall task intensity and a large number of theoretical maintenance times, thanks to the distributed maintenance task allocation heuristic rule, the allocation and transfer of faulty maintenance tasks and the replenishment of available carrier-based aircraft are realized, so that the platform can always meet the fleet availability required for the mission and ensure the stable and full completion of the mission wave, which is manifested as an advantage in the WR indicator. 3. The stable task completion and wave rotation continuity indicators brought about by the advantages of the cross-platform maintenance linkage mechanism in the CR indicator ensure the overall stability of combat training tasks and deck operations. 4. Thanks to the advantages of the sub-indicators, the cross-platform maintenance linkage mechanism has obtained the overall indicator compared with the single-platform independent maintenance. f The improvement is in line with the expected law. However, due to the high complexity and strong uncertainty of the scheduling process of cross-platform transfer of carrier-based aircraft, the discreteness of the solution results under some case platforms is greater than that of independent maintenance of a single platform, and the overall stability of the solution results is slightly weaker, which is manifested in some cases. f 5. The independent maintenance of a single platform causes the maintenance load of the platform hangar to be concentrated under the condition of heavy maintenance tasks, resulting in a backlog of maintenance tasks and poor task timeliness, which is manifested as a shortage of task waves, that is, the WR index is less than 1, making the index f Compared with MR, the cross-platform maintenance linkage mechanism shows disadvantages, but since the maintenance of a single platform avoids breaking the cumbersome and complicated process of breaking the dispatch cycle operation process of the original platform of some carrier-based aircraft and the transfer of maintenance tasks, the scheduling result is slightly more stable, which is manifested in some cases. f Advantages in variance and CR indicators.

[0187] Table 5

[0188]

[0189] Table 5 (continued)

[0190]

[0191] Experimental conclusion:

[0192] 1) Aiming at the problem of integrated scheduling of distributed platform fleet wave dispatch and maintenance, this paper systematically analyzes theoretical research on aircraft task assignment and maintenance planning, carrier-based aircraft dispatch and recovery operation scheduling, carrier-based aircraft maintenance scheduling, dynamic scheduling and other fields in the field of civil aviation at home and abroad. Taking the task wave completion rate, maintenance load dispersion and wave rotation continuity as optimization goals, and comprehensively considering various processes and constraints, a DP-ASRS dynamic model of maritime distributed platforms considering maintenance task transfer is established, and a two-stage optimization solution method is proposed, which can provide theoretical support for commanders to formulate fleet allocation, dispatch flight plans and collaborative maintenance scheduling plans for future distributed maritime operations.

[0193] 2) The established model can meet the constraints of various processes, resources and time for the allocation of distributed platform carrier-based aircraft fleet, flight sequence planning, cyclic dispatch operations, deck service support, and maintenance. The Gantt chart scheduling results verify the correctness and effectiveness of the model.

[0194] 3) The proposed two-stage optimization solution framework - the distributed platform fleet allocation agent function decoupling mechanism significantly improves the computing efficiency while ensuring the scheduling optimization performance, and realizes the rationalization and efficiency of the integrated scheduling plan for the wave dispatch and maintenance of the offshore distributed platform fleet.

[0195] 4) Compared with the independent maintenance of a single platform, the cross-platform maintenance linkage mechanism shows advantages in all indicators set by the model, especially under the conditions of high mission intensity, high probability of failure and high theoretical maintenance times. Therefore, the cross-platform maintenance linkage mechanism can be applied to actual combat and training tasks.

[0196] It should be understood that although Figure 5 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Figure 5 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequentially, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0197] See also Fig.15, a two-layer optimization decoupling device 100 based on a proxy function is also provided, comprising a data acquisition module 11, a model calling module 12 and an optimization output module 13. Among them, the data acquisition module 11 is used to obtain the current mission flight plan parameters and maintenance task parameters; the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft on the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations of cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single wave deck service support duration, the platform influence weight and the efficiency evaluation index influence weight, and the maintenance task parameters include the number of maintenance skills, the number of platform maintenance personnel and the maintenance mode. The model calling module 12 is used to call the constructed DP-ASRS dynamic model; the objective function of the DP-ASRS dynamic model includes: the distributed platform task comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function, and the constraints of the DP-ASRS dynamic model include the distributed platform carrier-based aircraft fleet allocation constraints, the flight timing planning constraints, the maintenance scheduling constraints and the distributed maintenance task platform transfer constraints. The optimization output module 13 is used to optimize and solve the DP-ASRS dynamic model according to the mission flight plan parameters and maintenance mission parameters using a two-stage optimization method, and decouple and output the current integrated scheduling plan; the integrated scheduling plan is used for integrated scheduling of distributed platform fleets.

[0198] The above-mentioned two-layer optimization decoupling device 100 based on the proxy function, through the collaboration of various modules, for the currently required integrated scheduling of the distributed platform fleet, first obtains the current mission flight plan parameters and maintenance mission parameters, and then calls the constructed DP-ASRS dynamic model, and uses the acquired known parameter data to optimize and solve the DP-ASRS dynamic model through a two-stage optimization method, decouples and outputs the distributed platform task comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function to maximize the optimal scheduling plan that meets the constraints of the model, that is, the current integrated scheduling plan is obtained, and finally the integrated scheduling plan can be used to carry out the integrated scheduling of the distributed platform fleet, thereby achieving the purpose of efficiently realizing the integrated scheduling of the distributed platform fleet, and having strong real-time and applicability.

[0199] In one embodiment, each module of the above-mentioned two-layer optimization decoupling device 100 based on proxy function can also be used to implement the corresponding sub-steps in each embodiment of the above-mentioned two-layer optimization decoupling method based on proxy function.

[0200] Regarding the specific limitations of the two-layer optimization decoupling device 100 based on proxy functions, please refer to the corresponding limitations of the two-layer optimization decoupling method based on proxy functions above, which will not be repeated here. Each module in the above-mentioned two-layer optimization decoupling device 100 based on proxy functions can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of a device with a specific data processing function in the form of hardware, or can be stored in the memory of the aforementioned device in the form of software, so that the processor can call and execute the operations corresponding to the above modules. The aforementioned device can be, but is not limited to, a data management / command device or a computing system for an offshore distributed platform.

[0201] On the other hand, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor can implement the following steps when executing the computer program: obtaining current mission flight plan parameters and maintenance task parameters; calling the constructed DP-ASRS dynamic model; optimizing and solving the DP-ASRS dynamic model according to the mission flight plan parameters and maintenance task parameters using a two-stage optimization method, and decoupling and outputting the current integrated scheduling plan; the integrated scheduling plan is used for integrated scheduling of distributed platform fleets. Among them, the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft carried by the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single wave deck service support duration, the platform influence weight and the efficiency evaluation index influence weight; the maintenance task parameters include the number of maintenance skills, the number of platform maintenance personnel and the maintenance mode; the objective function of the DP-ASRS dynamic model includes: the distributed platform task comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function; the constraints of the DP-ASRS dynamic model include the distributed platform carrier-based aircraft fleet allocation constraints, the flight timing planning constraints, the maintenance scheduling constraints and the distributed maintenance task platform transfer constraints.

[0202] In one embodiment, when the processor executes the computer program, the added steps or sub-steps in each embodiment of the above-mentioned two-layer optimization decoupling method based on the proxy function can also be implemented.

[0203] On the other hand, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: obtaining current mission flight plan parameters and maintenance task parameters; calling the constructed DP-ASRS dynamic model; optimizing and solving the DP-ASRS dynamic model according to the mission flight plan parameters and maintenance task parameters using a two-stage optimization method, and decoupling and outputting the current integrated scheduling plan; the integrated scheduling plan is used for integrated scheduling of distributed platform fleets. Among them, the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft carried by the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single wave deck service support duration, the platform influence weight and the efficiency evaluation index influence weight; the maintenance task parameters include the number of maintenance skills, the number of platform maintenance personnel and the maintenance mode; the objective function of the DP-ASRS dynamic model includes: the distributed platform task comprehensive efficiency optimization objective function and the single platform efficiency optimization objective function; the constraints of the DP-ASRS dynamic model include the distributed platform carrier-based aircraft fleet allocation constraints, the flight timing planning constraints, the maintenance scheduling constraints and the distributed maintenance task platform transfer constraints.

[0204] In one embodiment, when the computer program is executed by a processor, it can also implement the steps or sub-steps added in each embodiment of the above-mentioned two-layer optimization decoupling method based on the proxy function.

[0205] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus dynamic random access memory (Rambus DRAM, referred to as RDRAM) and interface dynamic random access memory (DRDRAM).

[0206] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0207] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and all of them belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A two-layer optimization decoupling method based on proxy function, characterized in that: Includes steps: Obtaining current mission flight plan parameters and maintenance mission parameters; the mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft on the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single-wave deck service support duration, the platform impact weight, and the performance evaluation index impact weight; the maintenance mission parameters include the number of maintenance skills, the number of platform maintenance personnel, and the maintenance mode; Call the constructed DP-ASRS dynamic model; The objective function of the DP-ASRS dynamic model includes: a distributed platform task comprehensive effectiveness optimization objective function and a single platform effectiveness optimization objective function, and the constraints of the DP-ASRS dynamic model include distributed platform carrier-based aircraft fleet allocation constraints, flight timing planning constraints, maintenance scheduling constraints, and distributed maintenance task platform transfer constraints; According to the mission flight plan parameters and maintenance mission parameters, the DP-ASRS dynamic model is optimized and solved by using a two-stage optimization method, and the current integrated scheduling scheme is decoupled and output; the integrated scheduling scheme is used for integrated scheduling of distributed platform fleets; Taking the constructed distributed platform cluster allocation proxy function as the objective function, the distributed platform cluster allocation result of the first stage optimization is obtained by using the evolutionary algorithm. The distributed platform cluster allocation result is used to perform a second-stage optimization operation; wherein the second-stage optimization operation includes: Input the distributed platform cluster allocation result to initialize the population; Decode each individual in the population until all task waves are completed; Calculate the fitness function value of the individual decoding generation scheme; where the fitness function is the single platform performance optimization target; Perform population genetic evolution operations; Repeat the steps of decoding each individual in the population until all task waves are completed, and then perform the population genetic evolution operation until the iteration end condition is met; Output the flight plan and performance optimal results of each platform; wherein the population genetic evolution operation includes: The population performs a binary tournament selection operation, selecting individuals with better fitness as parent individuals, generating a parent population of half the size of the population; Perform a two-point crossover operation between parent populations to generate offspring populations; The offspring population performs random key mutation operations; The distributed platform cluster allocation proxy function is: ； in, , and represents the influence weight of the three rules, Indicates p Each platform carries a collection of carrier-based aircraft. P Indicates the number of distributed platforms, Indicates p Platform i The theoretical maintenance time for a carrier-based aircraft to perform scheduled maintenance is Indicates p Platform i The theoretical maintenance time for a carrier-based aircraft to perform scheduled maintenance is express p The ratio of the average mission sorties of platform carrier-based aircraft to the average mission sorties of distributed platform carrier-based aircraft. express p The ratio of the number of platform maintenance resources configured to the total number of distributed platform maintenance resources, For the p The ratio of the total available time of each platform cluster before the next scheduled maintenance to the total available time of the distributed platform cluster before the next scheduled maintenance, For the p The ratio of the sum of the maximum theoretical maintenance time of a single platform cluster to the sum of the maximum theoretical maintenance time of the total distributed platform cluster. The number of carrier-based aircraft platforms, For the p The number of carrier-based aircraft carried by each platform, is a non-negative function, defined as: ; The objective function of the distributed platform task comprehensive performance optimization is: : ； The single platform performance optimization objective function is: : ； in, express p Platform performance evaluation value, Indicates the influence weight of each platform ， , and Indicates the impact weight of the three levels of performance indicators ，P Indicates the number of distributed platforms, represents the task wave completion rate, Indicates the maintenance load distribution, Indicates the continuity of wave rotation; The completion rate of the task wave is: ； in, is a task-complete wave variable if and only if hour, For the p The carrier-based aircraft of each platform are in waves Perform flight missions, Indicates w The wave is a task-completed wave; otherwise Indicates The wave is a vacant wave. Indicates p Each platform carries a collection of carrier-based aircraft. Indicates the number of carrier-based aircraft required for a single wave flight mission of the platform, Indicates p A set of platform mission dispatch waves, Indicates p The number of mission waves per platform; The maintenance load dispersion is: ； in, is the standard deviation calculation function, Indicates p Each platform's carrier-based aircraft is assembled for maintenance; The wave rotation continuity is: ； in, is a continuous rotating wave variable if and only if , When established, Indicates i Ship-borne aircraft Complete continuous rotation within the wave interval, otherwise Indicated in The continuous rotation is not completed within the wave interval; Indicates p Each platform carries a collection of carrier-based aircraft. Indicates p The platform task dispatch wave set is represented by: Indicates p The number of mission waves per platform, P Indicates the number of distributed platforms, Indicates p The number of continuous rotation waves of the platform cycle operation, Indicates p The number of wave groups dispatched by each platform in a cycle, For the p The number of carrier-based aircraft required for a single-wave flight mission for each platform; .

2. The two-layer optimization decoupling method based on proxy function according to claim 1 is characterized in that: The step of decoding each individual of the population until all task waves are completed includes: Determine whether the available cluster needs maintenance, if so, perform the first operation, otherwise perform the second operation; Execute the first operation; the first operation includes inputting the maintenance mode and the maintenance start time, judging whether to perform maintenance task platform transfer according to the distributed maintenance task allocation heuristic rule, and if so, executing the maintenance task platform transfer operation, otherwise judging whether to perform reactive rescheduling; wherein, if reactive rescheduling is performed, inputting the maintenance baseline scheduling result, otherwise initializing the maintenance task priority code according to the first-come-first-served rule, executing the maintenance task decoding operation, allocating maintenance personnel based on the constraint and load balancing principle, saving the maintenance baseline scheduling result, and outputting the maintenance completion time of each carrier-based aircraft; Execute the second operation; the second operation includes determining whether the single-wave flight mission is completed, determining whether the maintenance mission is completed, and updating the available fleet; Select carrier-based aircraft for single-wave sorties in accordance with flight sequence planning constraints; Return to the step of determining whether the available fleet needs maintenance, if so, perform the first operation, otherwise perform the second operation, until all task waves are completed.

3. The two-layer optimization decoupling method based on proxy function according to claim 2 is characterized in that: The distributed maintenance task allocation heuristic rule is: ； in, is the absolute value function, express p platform t Changes in the number of carrier-based aircraft to be repaired before and after the allocation of maintenance tasks at each moment, For the p Platforms t The number of carrier-based aircraft waiting for repair at any time, : No. p Platforms t The number of carrier-based aircraft waiting for maintenance after maintenance tasks are assigned at any time. express p platform t Always check resource utilization.

4. The two-layer optimization decoupling method based on proxy function according to claim 1 is characterized in that: The distributed platform carrier-based aircraft fleet allocation constraints include: the number of carrier-based aircraft allocated to the platform matches the number of carrier-based aircraft required for the platform mission and each carrier-based aircraft can only be allocated to one platform in the same allocation; The flight sequence planning constraints include: selection constraints of carrier aircraft wave rotation, adjacent wave formation and wave rotation, mutual exclusion constraints of carrier aircraft execution mission and maintenance / residence, constraints on the number of carrier aircraft dispatched for a single wave flight mission, constraints on the timing of carrier aircraft performing regular maintenance, and constraints on the association of flight sequence planning and maintenance scheduling; The maintenance scheduling constraints include: each maintenance process is carried out in sequence according to the established maintenance process, and the subsequent maintenance task must start after the completion of the previous process; each maintenance process has guaranteed start time and guaranteed end time constraints; the demand for skills of each maintenance process matches the number of maintenance personnel assigned to each maintenance process and who choose this skill; any maintenance personnel can only use one skill when performing maintenance work on any maintenance process; maintenance personnel conflict priority constraints; The distributed maintenance task platform transfer constraints include: the number of maintenance to be transferred is less than or equal to the total number of carrier-based aircraft to be repaired; the total number of global transfers is equal to the total number of supplements.

5. A two-layer optimization decoupling device based on a proxy function for implementing the method of claim 1, characterized in that: include: Data acquisition module, used to obtain the current mission flight plan parameters and maintenance mission parameters; The mission flight plan parameters include the number of mission waves, the number of distributed platforms, the number of carrier-based aircraft on the platform, the number of continuous rotation waves of cyclic operations, the number of wave formations for cyclic sorties, the number of carrier-based aircraft required for a single wave, the flight cycle duration, the deck operation cycle duration, the single-wave deck service support duration, the platform impact weight, and the performance evaluation index impact weight; the maintenance mission parameters include the number of maintenance skills, the number of platform maintenance personnel, and the maintenance mode; Model calling module, used to call the constructed DP-ASRS dynamic model; The objective function of the DP-ASRS dynamic model includes: a distributed platform task comprehensive effectiveness optimization objective function and a single platform effectiveness optimization objective function, and the constraints of the DP-ASRS dynamic model include distributed platform carrier-based aircraft fleet allocation constraints, flight timing planning constraints, maintenance scheduling constraints, and distributed maintenance task platform transfer constraints; An optimization output module is used to optimize and solve the DP-ASRS dynamic model according to the mission flight plan parameters and the maintenance mission parameters using a two-stage optimization method, and decouple and output the current integrated scheduling plan; The integrated scheduling solution is used for integrated scheduling of distributed platform clusters.

6. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the two-layer optimization decoupling method based on the proxy function as described in any one of claims 1 to 4 when executing the computer program.

Images