CN118627845A - A method for scheduling aircraft offshore platform operations - Google Patents

Abstract

The present invention relates to an aircraft offshore platform operation scheduling method, belonging to the technical field of operation scheduling methods. The method comprises the following steps: 1. Setting a scheduling model and initialization parameters of the scheduling model; 2. Using multi-layer coding to initialize the population; 3. Using serial scheduling to decode the population, generating a scheduling timing plan and a resource allocation plan; 4. Selecting a parent individual population for the population initialized in step 2; 5. Updating the individuals in the parent individual population in step 4 to obtain a child individual population; 6. Using step 3 to decode the child individuals in the child individual population in step 5; 7. Calculating the comprehensive performance of the decoded child individuals in step 6 and retaining the historical optimal individuals; 8. Updating the coding of the population, repeating steps 5 to 7, and obtaining the scheduling plan corresponding to the optimal individual. The method can realize the timing planning of the aircraft group deck operation scheduling and the scientific allocation of related resources.

Description

A method for scheduling aircraft offshore platform operations

Technical Field

The invention relates to an aircraft offshore platform operation scheduling method, in particular to an aircraft offshore platform operation scheduling method oriented to integrated integration, belonging to the technical field of operation scheduling methods.

Background Art

Aircraft carriers and their battle groups are key forces in far sea operations. They can be deployed at the forefront of global waters, project forces, safeguard maritime rights and interests, and dominate modern maritime battlefields. The powerful combat capability of offshore platforms must be achieved through the deployment and recovery of aircraft, and efficient deck operations are the key to improving the deployment and recovery capabilities. As a representative of high-tech intensive military system engineering, aircraft deck operation scheduling involves the coordinated implementation of key operations such as storage and deck transfer, maintenance and service support, airborne weapons transfer, and dispatch and departure under limited deck space and equipment resources. It requires the overall allocation of resources and scientific arrangement of timing.

At present, the deck operation still cannot be dispatched as a whole, resulting in low comprehensive support efficiency and low resource utilization. In essence, first, the scheduling of deck operations is complex and diverse. The diversity of the operation process at each stage leads to different scheduling model characteristics, covering multiple types of problems such as single-machine scheduling, mixed workshop scheduling, resource-constrained multi-project scheduling, and vehicle path planning, forming a heterogeneous distributed scheduling system, which is difficult to unify modeling and integration; second, the inherent mechanism of the coupling mechanism is still unclear, and the coupling relationship structure between the operation scheduling systems at each stage is complex, showing a non-hierarchical coupling structure, which increases the difficulty of solving; third, there is a lack of global performance indicators, including a series of performance indicators such as operation time, resource load balancing, and resource utilization. It is difficult to consider them globally, which limits further integration, expansion, and optimization.

Summary of the invention

The purpose of the present invention is to solve the problems in the prior art and to provide a method for scheduling aircraft offshore platform operations. The method can realize the timing planning of aircraft group deck operation scheduling and the scientific allocation of related resources, thereby improving the comprehensive support efficiency and resource utilization of the aircraft group, and improving the response speed and quality of scheduling command decisions.

In order to solve the above problems, the present application is implemented through the following technical solutions:

A method for scheduling aircraft offshore platform operations, which is special in that it includes the following steps:

S1. Setting the scheduling model and initialization parameters of the scheduling model: Select the simulation object, set the number of simulation objects, and configure the corresponding supporting settings to establish a scheduling model. The scheduling model can complete the integrated support of the ship deck including storage and deck transfer, maintenance and service support, and dispatch and departure;

S2, use multi-layer coding to initialize the population;

S3, using serial scheduling to decode the population and generate a scheduling timing plan and a resource allocation plan;

S4, selecting a parent individual population for the population initialized in step S2;

S5, updating the individuals in the parent individual population in step S4 to obtain the offspring individual population;

S6, decoding the offspring individuals in the offspring individual population in step S5 using step S3;

S7, calculating the comprehensive performance of the offspring individuals decoded in step S6 and retaining the best individual in history;

S8. Update the population code, repeat steps S5 to S7, and obtain the scheduling solution corresponding to the optimal individual.

Furthermore, the step S1 is specifically as follows: taking the offshore platform fleet as the simulation object, setting the fleet type to fixed-wing aircraft of the same model, setting the combat target to execute the opposite combat mission, the fleet dispatch scale is n aircraft dispatch, and n aircraft need to be transferred from the hangar to the deck parking position to complete the ship-deck integrated support including storage and deck transfer, maintenance service support, and dispatch and departure; the supporting settings include maintenance support procedures, maintenance support personnel, service support equipment and bomb hanging process settings.

Furthermore, step S2 includes a four-layer coding mechanism according to the characteristics of different stages of aircraft integrated collaborative scheduling, specifically:

The first layer: priority coding for aircraft transfer, using real number coding, let the coding be expressed as , where any dimension gene is a random real number in the interval (0, 1), The number of transfer aircraft;

The second layer: ensure the parking space layout coding, ensure the parking space layout uses integer coding, , where any dimension gene represents the parking stand number of the i-th aircraft, Number of aircraft to be supported on deck:

The third level: Maintenance service support process priority coding, using priority coding based on the support process correction, the start time of the process Convert to a coding priority sequence, let , where genes represents the priority of the jth process of the i-th aircraft, which is corrected to the start time of the scheduling work;

The fourth layer: the departure phase coding, using real number coding, set is the number of takeoff positions, let the encoding be , where any dimension gene For interval The integer part Int( _xi ) represents the take-off position number of the i-th aircraft, and the decimal part Dec( _xi ) represents the priority order number of the i-th aircraft. The number of aircraft waiting to be supported on the deck.

Furthermore, the step S3 performs serial decoding on the coding of step S2. The decoding process is the process of converting the priorities of the above-mentioned aircraft/processes into a scheduling plan, and serial decoding is performed for the aircraft transfer, maintenance service support, and dispatch and departure stages.

Furthermore, the decoding process of the aircraft transfer scheduling is as follows:

Step 1.1: Assume that the set of aircraft to be scheduled is ,initialization , the set of scheduled aircraft is ,initialization , ;

Step 1.2: When , go to step 1.5;

Step 1.3: The aircraft number currently waiting for transfer First, query the aircraft set of the initial layout state corresponding to the interfering parking positions in each stage Whether it has been scheduled. If it has been scheduled, the latest transfer time of the aircraft at the previous intervening parking position will be used as the transfer start time of the aircraft to be scheduled. ;

If it is not dispatched, the current transfer path is blocked and the current aircraft cannot be transferred. , go to step 1.2;

Next, check the scheduled aircraft Whether the parking space of the aircraft currently waiting to be dispatched blocks the transfer path, and the time of entering the space is earlier than , if it exists, it cannot be transferred, and the same order , go to step 1.2;

Step 1.4: After the current aircraft is successfully transferred, update the status parameters. , , and at the same time , go to step 1.2 loop iteration judgment;

Step 1.5: After all current aircraft have completed their transfer missions, the scheduling schedule is entered into step 2.1.

Furthermore, the steps of decoding in the maintenance service support stage are as follows:

Step 2.1. Definition is the jth process of the i-th aircraft, the entry time of aircraft i is Ex _i , let , , for resource status initialization, let ;

Step 2.2: Go to step 2.6, where: For the completed maintenance service support process set, is the set of processes being executed;

Step 2.3: According to the set of processes being executed Calculate the earliest completion time , and the process will end Merge into the completed maintenance support process set , query the set of unexecuted candidate scheduling processes that currently meet the constraints , go to step 2.4;

Step 2.4: Scheduled process set never executed Select the highest priority process to perform maintenance support, allocate resources and personnel according to the process priority, update the start and completion time of the support operation, update the equipment occupancy status, and Add to , until ;

Step 2.5, command , go to step 2.2;

Step 2.6: After all current aircraft have completed maintenance services, the updated scheduling schedule is input into step 2.1.

Furthermore, the decoding for the aircraft departure phase uses the completion time of the maintenance service support phase as the start time of the aircraft departure, based on the coding of the parking stand arrangement status, the scheduling takeoff position and the dispatch priority, and the specific decoding process is as follows:

Step 3.1: Define the set of aircraft to be dispatched , aircraft assembly has been dispatched , the aircraft is currently completing the mission phase , the number of operation stages completed by the i-th aircraft is ,initialization , , , ;

Step 3.2: When , go to step 3.4;

Step 3.3: According to the number of the aircraft to be dispatched , the take-off position corresponding to the aircraft to be dispatched , query the current completed job stage and then execute the next stage of the job, ;

Step 3.3.1, check the warm-up position status: first determine whether the current parking position is an available warm-up position, if so, go to step 5, and at the same time If the warm-up position is not available, it is necessary to determine whether there is an empty warm-up position. If there is a vacancy, it is transferred according to the principle of proximity, that is, the transfer time is the shortest. , go to step 5; if there is no vacant warm-up position, continue to determine whether the aircraft has been dispatched to assemble Is it an empty set? If not, select the earliest scheduling time and parking position from the scheduled aircraft set for warm-up. , go to step 3.3.2; if all the conditions are not met, warm-up cannot be achieved, let , go to step 3.2;

Step 3.3.2: Check whether there is any interference in the dispatch path from the warm-up position to the take-off position: Assume that the warm-up position reaches the target take-off position The interfering parking positions on the scheduling path Is there any aircraft stationed and whether it is dispatched? If there are unscheduled aircraft stationed, the dispatch conditions are not met. , go to step 2; if there is an aircraft scheduled to park at an interfering parking stand, the scheduling time of the interfering parking stand is used as the earliest taxi departure time for the aircraft , , go to step 3.3.3;

Step 3.3.3: Since the aircraft needs to wait for the deflector plate to cool and reset before it can enter the position for deployment preparation, the previous aircraft release time plus the deflector plate reset time As the aircraft enters the parking position, , go to step 3.3.4;

Step 3.3.4, determine the wake turbulence interval time: take the previous aircraft release time plus the wake turbulence interval time as the earliest release time t of the aircraft, and let , , , go to step 3.2 to continue the iteration;

Step 3.4: Output the aircraft integrated scheduling timing allocation plan.

Furthermore, the step S4 adopts a binary tournament strategy to select the parent population, which is as follows:

S41. Determine the number of individuals to be selected each time: In the binary tournament selection, two individuals are selected for comparison each time;

S42, randomly select individuals: randomly select 50 individuals from the population initialized in step S2;

S43, selecting individuals according to fitness values: according to the total time used by the scheduling scheme of each individual, selecting the individual with the shortest total time used by the scheduling scheme to enter the next generation population;

S44, repeat the selection process: repeat the above steps to the maximum number of evaluations, until the size of the parent population reaches the size of the initialization population; the fitness value refers to the total time used for the scheduling plan.

Furthermore, the step S5 uses crossover and mutation operations to update the individuals in the parent population of step S4, as follows:

Crossover operation: For the three-level coding of transfer priority, maintenance and service support process, and departure, two-point crossover is performed by randomly selecting two intersection points; for the coding of parking space layout, single-point crossover is adopted. If there are duplicate genes after crossover, the duplicate genes are deleted and selected again from the available optional parking spaces;

Mutation operation: For the three-level codes of transfer priority, maintenance and service support procedures, and dispatch and departure, random key mutation is performed with a random probability of 5%; for the parking bay layout code, in order to meet the extension update and internal optimization of the gene search domain, a 50% probability of performing random key mutation and a 50% probability of performing exchange mutation are set;

Two strategies are used to update individual genes and expand the search range of solutions:

Individual exchange strategy: synchronously select two aircraft within an individual to exchange gene fragments of the same process, select an individual gene to exchange the priority of the same process of different aircraft, and form a new individual gene;

Individual recombination strategy: randomly select an aircraft within the individual to reorganize the gene fragments of the process, select an individual gene to reorganize the priorities of different processes of the same carrier-based aircraft, and form a new individual gene.

Furthermore, the offspring individuals in step S7 calculate the comprehensive performance The details are as follows: , where: Cmax is the historical maximum completion time, Cmin is the historical minimum completion time, and C is the completion time of the offspring individual calculated this time;

After calculating the comprehensive performance, the individual with the shortest total time used in the scheduling scheme is selected and compared with the parent individual with the shortest total time used in the scheduling scheme in step S4, and the historical best individual is retained.

Furthermore, the specific steps of step S8 are as follows: after retaining the historical best individual in step S7, replace the individual with the lowest fitness in the population with the best individual to form a new individual coding population, repeat steps S5 to S7 for the new individual coding population until the number of decoding evaluations meets the iterative termination condition of the maximum number of evaluations, then end the decoding and output the scheduling plan corresponding to the historical best individual.

Furthermore, the maximum number of evaluations in the S44, repeated selection process and step S8 is 100.

The aircraft offshore platform operation scheduling method of the present application is faster than manual experience scheduling in terms of simulation time; simulation data shows that it effectively improves the fleet dispatch rate by more than 5% and reduces the labor intensity of operators by more than 10%; in terms of economic benefits, it reduces task planning and execution time, further promotes reduction in staff and increase in efficiency, and reduces onboard resource consumption.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a flow chart of the present invention;

FIG2 is a decoding flow chart of FIG1 ;

Figure 3 is a single machine assurance process AON diagram;

Figure 4 is a Gantt chart of integrated scheduling of aircraft groups;

Figure 5 is a Gantt chart of the transport equipment;

Figure 6 is a Gantt chart for the bomb hanging group scheduling;

Figure 7 shows the guarantee performance convergence curve.

DETAILED DESCRIPTION

The following is a specific implementation of the present invention with reference to the accompanying drawings to further illustrate the composition of the present invention.

Embodiment 1. A method for scheduling aircraft offshore platform operations, the process of which is shown in FIG1 , comprises the following steps:

S1. Setting the scheduling model and initialization parameters of the scheduling model: Select the simulation object, set the number of simulation objects, and configure the corresponding supporting settings to establish a scheduling model. The scheduling model can complete the integrated support of the ship deck including storage and deck transfer, maintenance and service support, and dispatch and departure;

S2, use multi-layer coding to initialize the population;

S3, using serial scheduling to decode the population and generate a scheduling timing plan and a resource allocation plan;

S4, selecting a parent individual population for the population initialized in step S2;

S5, updating the individuals in the parent individual population in step S4 to obtain the offspring individual population;

S6, decoding the offspring individuals in the offspring individual population in step S5 using step S3;

S7, calculating the comprehensive performance of the offspring individuals decoded in step S6 and retaining the best individual in history;

S8. Update the population code, repeat steps S5 to S7, and obtain the scheduling solution corresponding to the optimal individual.

Furthermore, the step S1 is specifically as follows: taking the offshore platform fleet as the simulation object, setting the fleet type to fixed-wing aircraft of the same model, setting the combat target to execute the opposite combat mission, the fleet dispatch scale is n aircraft dispatch, and n aircraft need to be transferred from the hangar to the deck parking position to complete the ship-deck integrated support including storage and deck transfer, maintenance service support, and dispatch and departure; the supporting settings include maintenance support procedures, maintenance support personnel, service support equipment and bomb hanging process settings.

Furthermore, step S2 includes a four-layer coding mechanism according to the characteristics of different stages of aircraft integrated collaborative scheduling, specifically:

The first layer: priority coding for aircraft transfer, using real number coding, let the coding be expressed as , where any dimension gene For interval A random real number on , The number of transfer aircraft;

The second level: the parking stand layout coding. Since the aircraft support adopts a centralized support mode, that is, all maintenance support and dispatch preparation work are completed at the same maintenance station, and the aircraft is driven out in place, and only one traction vehicle is needed for the whole process. Therefore, the parking stand layout is also the pre-dispatch layout state. The parking stand layout adopts integer coding, so , where any dimension gene represents the parking stand number of the i-th aircraft, Number of aircraft to be supported on deck:

The third level: Maintenance service support process priority coding, using priority coding based on the support process correction, the start time of the process Convert to a coding priority sequence, let , where genes represents the priority of the jth process of the i-th aircraft, which is corrected to the start time of the scheduling work;

The fourth level: the coding of the departure phase. Since there are few aircraft constrained by the warm-up position when the aircraft is dispatched, most aircraft can be warmed up at the original position. Those that cannot be warmed up at the original position can be deployed to the nearest warm-up position. Therefore, the departure is mainly selected based on the take-off position. Real number coding is used. is the number of takeoff positions, let the encoding be , where any dimension gene For interval The integer part Int( _xi ) represents the take-off position number of the i-th aircraft, and the decimal part Dec( _xi ) represents the priority order number of the i-th aircraft. The number of aircraft waiting to be supported on the deck.

A complete set of four-layer codes corresponds to a complete individual, that is, a complete scheduling plan, which is a feasible solution to the problem. Therefore, the above codes need to be decoded in series. The decoding process is the process of converting the priority of the above aircraft/processes into a scheduling plan, and serial decoding is performed for aircraft transfer, maintenance service support, and dispatch and departure stages.

Furthermore, the step S3 performs serial decoding on the code of step S2. The decoding process is a process of converting the priorities of the above-mentioned aircraft/processes into a scheduling plan, and serial decoding is performed for the aircraft transfer, maintenance service support, and departure stages;

In the transfer stage, if the existing aircraft arrangement hinders the aircraft transfer path, an infeasible scheduling solution is generated. Therefore, the transfer priority is the key to aircraft scheduling. The decoding process of the aircraft transfer scheduling is as follows:

Step 1.1: Assume that the set of aircraft to be scheduled is ,initialization ; The scheduled aircraft set is ,initialization , ;

Step 1.2: When , go to step 1.5;

Step 1.3: The aircraft number currently waiting for transfer First, query the aircraft set of the initial layout state corresponding to the interfering parking positions in each stage Whether it has been scheduled. If it has been scheduled, the latest transfer time of the aircraft at the previous intervening parking position will be used as the transfer start time of the aircraft to be scheduled. If not dispatched, the current transfer path is blocked and the current aircraft cannot be transferred. , go to step 1.2;

Next, check the scheduled aircraft Whether the parking space of the aircraft currently waiting to be dispatched blocks the transfer path, and the time of entering the space is earlier than , if it exists, it cannot be transferred, and the same order , go to step 1.2;

Step 1.4: After the current aircraft is successfully transferred, update the status parameters. , , and at the same time , go to step 1.2 loop iteration judgment;

Step 1.5: After all current aircraft have completed the transfer mission, the scheduling schedule is input into step 2.1;

For the decoding of the maintenance service support stage, the time when the transfer stage arrives at the target parking position is taken as the starting time of the maintenance service support stage, and the maintenance service support stage of the fleet is decoded based on the parking position layout code;

Definition: Completed maintenance service support process set , the set of candidate scheduling processes that have not been executed and the set of operations being scheduled The basic process of decoding is to select the process set to be scheduled in any scheduling period. , time parallel advancement records the earliest end time t of the process, and the set of scheduling processes has never been executed Query the processes that meet the scheduling conditions and complete the scheduling according to the priority of the processes;

The steps of decoding in the maintenance service support stage are as follows:

Step 2.1. Definition is the jth process of the i-th aircraft, the entry time of aircraft i is Ex _i , let , , for resource status initialization, let ;

Step 2.2: Go to step 2.6, where: For the completed maintenance service support process set, is the set of processes being executed;

Step 2.3: According to the set of processes being executed Calculate the earliest completion time , and the process will end Merge into the completed maintenance support process set , query the set of unexecuted candidate scheduling processes that currently meet the constraints , go to step 2.4;

Step 2.4: The scheduling process set has never been executed Select the highest priority process to perform maintenance support, allocate resources and personnel according to the process priority, update the start and completion time of the support operation, update the equipment occupancy status, and Add to , until ;

Step 2.5: , go to step 2.2;

Step 2.6: After all current aircraft have completed maintenance services, the updated scheduling schedule is input into step 2.1;

The decoding for the aircraft departure phase uses the completion time of the maintenance service support phase as the start time of the aircraft departure, based on the coding of the parking stand arrangement status and the scheduling takeoff position and the departure priority. The process is shown in Figure 2, and the specific decoding process is as follows:

Step 3.1: Define the set of aircraft to be dispatched , aircraft assembly has been dispatched , the aircraft is currently completing the mission phase , the number of operation stages completed by the i-th aircraft is ,initialization , , , ;

Step 3.2: When , go to step 3.4;

Step 3.3: According to the number of the aircraft to be dispatched , the take-off position corresponding to the aircraft to be dispatched , query the current completed job stage and then execute the next stage of the job, ;

Step 3.3.1, check the warm-up position status: first determine whether the current parking position is an available warm-up position, if so, go to step 5, and at the same time If the warm-up position is not available, it is necessary to determine whether there is an empty warm-up position. If there is a vacancy, it is transferred according to the principle of proximity, that is, the transfer time is the shortest. , go to step 5; if there is no vacant warm-up position, continue to determine whether the aircraft has been dispatched to assemble Is it an empty set? If not, select the earliest scheduling time and parking position from the scheduled aircraft set for warm-up. , go to step 3.3.2; if all the conditions are not met, warm-up cannot be achieved, let , go to step 3.2;

Step 3.3.2: Check whether there is any interference in the dispatch path from the warm-up position to the take-off position: Assume that the warm-up position reaches the target take-off position The interfering parking positions on the scheduling path Is there any aircraft stationed and whether it is dispatched? If there are unscheduled aircraft stationed, the dispatch conditions are not met. , go to step 2; if there is an aircraft scheduled to park at an interfering parking stand, the scheduling time of the interfering parking stand is used as the earliest taxi departure time for the aircraft , , go to step 3.3.3;

Step 3.3.3: Since the aircraft needs to wait for the deflector plate to cool and reset before it can enter the position for deployment preparation, the previous aircraft release time plus the deflector plate reset time As the aircraft enters the parking position, , go to step 3.3.4;

Step 3.3.4, determine the wake turbulence interval time: take the previous aircraft release time plus the wake turbulence interval time as the earliest release time t of the aircraft, and let , , , go to step 3.2 to continue the iteration;

Step 3.4: Output the aircraft integrated scheduling timing allocation plan.

Furthermore, the step S4 adopts a binary tournament strategy to select the parent population, which is as follows:

S41. Determine the number of individuals to be selected each time: In the binary tournament selection, two individuals are selected for comparison each time;

S42, randomly select individuals: randomly select 50 individuals from the population initialized in step S2;

S43, selecting individuals according to fitness values: according to the total time used by the scheduling scheme of each individual, selecting the individual with the shortest total time used by the scheduling scheme to enter the next generation population;

S44, repeat the selection process: repeat the above steps to the maximum number of evaluations, until the size of the parent population reaches the size of the initialization population; the fitness value refers to the total time used for the scheduling plan.

Furthermore, the step S5 uses crossover and mutation operations to update the individuals in the parent population of step S4, as follows:

Crossover operation: For the three-level coding of transfer priority, maintenance and service support process, and departure, two-point crossover is performed by randomly selecting two intersection points; for the coding of parking space layout, single-point crossover is adopted. If there are duplicate genes after crossover, the duplicate genes are deleted and selected again from the available optional parking spaces;

Mutation operation: For the three-level codes of transfer priority, maintenance and service support procedures, and dispatch and departure, random key mutation is performed with a random probability of 5%; for the parking bay layout code, in order to meet the extension update and internal optimization of the gene search domain, a 50% probability of performing random key mutation and a 50% probability of performing exchange mutation are set;

Two strategies are used to update individual genes and expand the search range of solutions:

Individual exchange strategy: synchronously select two aircraft within an individual to exchange gene fragments of the same process, select an individual gene to exchange the priority of the same process of different aircraft, and form a new individual gene;

Individual recombination strategy: randomly select an aircraft within the individual to reorganize the gene fragments of the process, select an individual gene to reorganize the priorities of different processes of the same carrier-based aircraft, and form a new individual gene.

Furthermore, the offspring individuals in step S7 calculate the comprehensive performance The details are as follows: , where: Cmax is the historical maximum completion time, Cmin is the historical minimum completion time, and C is the completion time of the offspring individual calculated this time;

After calculating the comprehensive performance, the individual with the shortest total time used in the scheduling scheme is selected and compared with the parent individual with the shortest total time used in the scheduling scheme in step S4, and the historical best individual is retained.

Furthermore, the specific steps of step S8 are as follows: after retaining the historical best individual in step S7, replace the individual with the lowest fitness in the population with the best individual to form a new individual coding population, repeat steps S5 to S7 for the new individual coding population until the number of decoding evaluations meets the iterative termination condition of the maximum number of evaluations, then end the decoding and output the scheduling plan corresponding to the historical best individual; the fitness value refers to the total time used for the scheduling plan.

Furthermore, the maximum number of evaluations in the S44, repeated selection process and step S8 is 100.

Embodiment 2. This embodiment is a simulation of the method described in Embodiment 1.

The offshore platform fleet is taken as the simulation object, and the fleet type is set as fixed-wing aircraft of the same type. The combat goal is set to perform opposite combat tasks. The fleet dispatch scale is 12 aircraft, and 12 aircraft need to be transferred from the hangar to the deck parking position to complete the ship-deck integrated support including in and out of the hangar and deck transfer, maintenance service support, and dispatch and departure; the aircraft assigned to the dispatch mission and the initial parking position are shown in Table 1.

Table 1: Initial parking positions for offshore platform aircraft

;

In terms of transportation equipment configuration, there are 4 hangar tractors, numbered ; There are 4 deck transfer vehicles, numbered ; Two lifts, numbered When performing transfer tasks, the transfer mooring and unmooring time is set to 30 seconds, the one-way operation time of the elevator is 70 seconds, and the upper limit of the hangar/deck tractor operating speed is 5km/h.

The supporting settings include maintenance support procedures, maintenance support personnel, service support equipment and bomb hanging process settings: the maintenance support process settings: because it is the same type of air-to-surface combat aircraft, the bombs are set to two air-to-air missiles and two surface missiles, and the AON diagram of the single-machine support process is constructed as shown in Figure 3. The marked processes in the figure have a demand for support configuration of 1; the maintenance support personnel settings: the transfer speed is 5km/h, according to the requirements of the n-machine combat mission, four majors and 5 special , 6 avionics , 12 ordnance , 17 mechanical A total of 40 maintenance personnel;

There are 7 categories of service support equipment, namely, gas station/cart, power supply station/cart, oxygen filling station/cart, nitrogen filling station\cart, hydraulic filling station\cart, power station/cart, and bomb hanging car, among which the power station/cart is a shared service support equipment, and the rest are exclusive equipment; the bomb hanging car is a mobile support equipment; the operation content of each process and the required maintenance support personnel profession and service support equipment are shown in Table 2.

Table 2: Table of content of guarantee process operation

;

The gas station/car, power supply station/car, oxygen filling station/car, nitrogen filling station\car, hydraulic filling station\car are numbered in sequence as follows: , the transfer speed is 3km/h, and the supply resources corresponding to each service equipment are , the original parking positions A2~A13 and B1~B4 are renumbered as 1-16, and the parking position range covered by their service support equipment is shown in Table 3;

Table 3: Parking space coverage of service support equipment

;

The take-off and departure settings: Considering the impact of warm-up on the island, A3-A6 parking positions cannot perform warm-up tasks; in terms of take-off position interference, since runway 1# and runway 3# overlap, C1 take-off position will interfere with runway 3# when performing take-off or waiting position preparation, A12 parking position and C3 waiting position interfere with each other, and the aircraft configuration performing the sortie mission meets the take-off position requirements;

The bomb hanging process setting is as follows: the air-to-air missile bomb hanging task is set to 1 and 2 processes, and the air-to-surface missile bomb hanging task is set to 3 and 4 processes; the bomb hanging group is divided into a manual bomb hanging group and a mechanical bomb hanging group. There are 7 manual bomb hanging groups, and the carrying limit for air-to-air and air-to-surface missiles is 4 and 2, and the corresponding bomb hanging time is 4 minutes and 7 minutes. There are 3 mechanical bomb hanging groups, and the carrying limit for air-to-air and air-to-surface missiles is 1, and the corresponding bomb hanging time is 4 minutes and 5 minutes. The transfer time is 5 km/h.

Figure 4 is an aircraft dispatch Gantt chart, covering the storage and deck transfer, maintenance support and departure phases. The horizontal axis is time, the vertical axis is the aircraft number, the white Gantt bar represents the station waiting or station transfer time, and the gray Gantt bar represents the mooring and unmooring time.

Taking aircraft I ₁ as an example, the initial position is at the hangar G4 position, and it is towed by the hangar tractor HM-2 to the elevator EM-1, and after rising to the ship deck, it is towed by the deck transfer vehicle DM-2 to the A6 position for maintenance and service support. Since the A6 position cannot be directly warmed up, the aircraft I ₁ needs to transfer to the A2 position to complete the warm-up self-check, and after waiting for a period of time, it is transferred from the A2 position to the take-off position C2 to complete the take-off and departure operations.

Figure 5 is a Gantt chart of the transport equipment scheduling, covering 4 hangar tractors , 4 deck transfer vehicles and two lifts The white Gantt bar in the figure represents the transfer time of the dispatched equipment, and the striped Gantt bar represents the time for the aircraft to be moored/unmoored.

Figure 6 is a Gantt chart of the bomb hanging group scheduling. The vertical axis is the hanging group number, including 7 human bomb hanging groups and 3 mechanical bomb hanging groups. The numbers in the figure are the support procedures that the aircraft needs to hang bombs, and the gray squares are the transfer time of support personnel/support equipment.

Figure 7 is the comprehensive performance iteration curve of the optimal individual of the population that is continuously evolving and iterating under the algorithm selection crossover mutation.

The simulation time shows that it is faster than manual experience scheduling; the simulation data shows that it can effectively increase the fleet dispatch rate by more than 5% and reduce the labor intensity of operators by more than 10%.

The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the defined scope.

Claims (10)

1. A method for scheduling aircraft offshore platform operations, characterized in that it includes the following steps: S1. Set the scheduling model and its initialization parameters: Select the simulation object, set the number of simulation objects, and configure the corresponding supporting settings to establish a scheduling model that can complete the integrated support of the ship deck including storage and deck transfer, maintenance and service support, and dispatch and departure; S2, use multi-layer coding to initialize the population; S3, using serial scheduling to decode the population and generate a scheduling timing plan and a resource allocation plan; S4, selecting a parent individual population for the population initialized in step S2; S5, updating the individuals in the parent individual population in step S4 to obtain the offspring individual population; S6, decoding the offspring individuals in the offspring individual population in step S5 using step S3; S7, calculating the comprehensive performance of the offspring individuals decoded in step S6 and retaining the best individual in history; S8. Update the population code, repeat steps S5 to S7, and obtain the scheduling solution corresponding to the optimal individual. 2. The method for scheduling aircraft offshore platform operations according to claim 1 is characterized in that: the step S1 specifically comprises: taking the offshore platform fleet as the simulation object, setting the fleet type to fixed-wing aircraft of the same model, setting the combat target to execute the opposite combat mission, and dispatching the fleet on a scale of n aircraft, requiring n aircraft to be transferred from the hangar to the deck parking position to complete the integrated ship-deck support including storage and deck transfer, maintenance service support, and dispatch and departure; the supporting settings include maintenance support procedures, maintenance support personnel, service support equipment, and bomb hanging procedure settings. 3. The method for scheduling an aircraft offshore platform operation according to claim 1 or 2, characterized in that: step S2 includes a four-layer coding mechanism according to the characteristics of different stages of aircraft integrated collaborative scheduling, specifically: The first layer: priority coding for aircraft transfer, using real number coding, let the coding be expressed as , where any dimension gene For interval A random real number on , The number of transfer aircraft; The second layer: ensure the parking space layout coding, ensure the parking space layout uses integer coding, , where any dimension gene represents the parking stand number of the i-th aircraft, Number of aircraft to be supported on deck: The third level: Maintenance service support process priority coding, using priority coding based on the support process correction, the start time of the process Convert to a coding priority sequence, let , where genes represents the priority of the jth process of the i-th aircraft, which is corrected to the start time of the scheduling work; The fourth layer: the departure phase coding, using real number coding, set is the number of takeoff positions, let the encoding be , where any dimension gene For interval The integer part Int( _xi ) represents the take-off position number of the i-th aircraft, and the decimal part Dec( _xi ) represents the priority order number of the i-th aircraft. The number of aircraft waiting to be supported on the deck. 4. The method for scheduling aircraft offshore platform operations according to claim 3 is characterized in that: step S3 performs serial decoding on the code of step S2, and the decoding process is the process of converting the priority of the aircraft/process into a scheduling plan, and serial decoding is performed for the aircraft transfer, maintenance service support, and departure stages. 5. The method for scheduling an aircraft offshore platform operation according to claim 4, characterized in that the decoding process of the aircraft transfer scheduling is as follows: Step 1.1: Assume that the set of aircraft to be scheduled is ,initialization , the set of scheduled aircraft is ,initialization , ; Step 1.2: When , go to step 1.5; Step 1.3: The aircraft number currently waiting for transfer First, query the aircraft set of the initial layout state corresponding to the interfering parking positions in each stage Whether it has been scheduled. If it has been scheduled, the latest transfer time of the aircraft at the previous intervening parking position will be used as the transfer start time of the aircraft to be scheduled. ; If it is not dispatched, the current transfer path is blocked and the current aircraft cannot be transferred. , go to step 1.2; Next, check the scheduled aircraft Whether the parking space of the aircraft currently waiting to be dispatched blocks the transfer path, and the time of entering the space is earlier than , if it exists, it cannot be transferred, and the same order , go to step 1.2; Step 1.4: After the current aircraft is successfully transferred, update the status parameters. , , and at the same time , go to step 1.2 loop iteration judgment; Step 1.5: After all current aircraft have completed their transfer missions, the scheduling schedule is entered into step 2.1. 6. The method for scheduling aircraft offshore platform operations according to claim 5, characterized in that the steps of decoding the maintenance service support stage are as follows: Step 2.1. Definition is the jth process of the i-th aircraft, the entry time of aircraft i is Ex _i , let , , for resource status initialization, let ; Step 2.2: Go to step 2.6, where: For the completed maintenance service support process set, is the set of processes being executed; Step 2.3: According to the set of processes being executed Calculate the earliest completion time , and the process will end Merge into the completed maintenance support process set , query the set of unexecuted candidate scheduling processes that currently meet the constraints , go to step 2.4; Step 2.4: The scheduling process set has never been executed Select the highest priority process to perform maintenance support, allocate resources and personnel according to the process priority, update the start and completion time of the support operation, update the equipment occupancy status, and Add to , until ; Step 2.5, command , go to step 2.2; Step 2.6: After all current aircraft have completed maintenance services, the updated scheduling schedule is input into step 2.1. 7. The method for scheduling aircraft offshore platform operations according to claim 6 is characterized in that: the decoding for the aircraft departure phase uses the completion time of the maintenance service support phase as the start time of the aircraft departure, based on the coding of the parking position arrangement status and the scheduling take-off position and the dispatch priority, and the specific decoding process is as follows: Step 3.1: Define the set of aircraft to be dispatched , aircraft assembly has been dispatched , the aircraft is currently completing the mission phase , the number of operation stages completed by the i-th aircraft is ,initialization , , , ; Step 3.2: When , go to step 3.4; Step 3.3: According to the number of the aircraft to be dispatched , the take-off position corresponding to the aircraft to be dispatched , query the current completed job stage and then execute the next stage of the job, ; Step 3.3.1, Query the warm-up position status: First determine whether the current parking position is an available warm-up position. If so, proceed to step 5 and If the warm-up position is not available, it is necessary to determine whether there is an empty warm-up position. If there is a vacancy, it is transferred according to the principle of proximity, that is, the transfer time is the shortest. , go to step 5; if there is no vacant warm-up position, continue to determine whether the aircraft has been dispatched to assemble Is it an empty set? If not, select the earliest scheduling time and parking position from the scheduled aircraft set for warm-up. , go to step 3.3.2; if all the conditions are not met, warm-up cannot be achieved, let , go to step 3.2; Step 3.3.2: Check whether there is any interference in the dispatch path from the warm-up position to the take-off position: Assume that the warm-up position reaches the target take-off position The interfering parking positions on the scheduling path Is there any aircraft stationed and whether it is dispatched? If there are unscheduled aircraft stationed, the dispatch conditions are not met. , go to step 2; if there is an aircraft scheduled to park at an interfering parking stand, the scheduling time of the interfering parking stand is used as the earliest taxi departure time for the aircraft , , go to step 3.3.3; Step 3.3.3: Since the aircraft needs to wait for the deflector plate to cool and reset before it can enter the position for deployment preparation, the previous aircraft release time plus the deflector plate reset time As the aircraft enters the parking position, , go to step 3.3.4; Step 3.3.4, determine the wake turbulence interval time: the previous aircraft release time plus the wake turbulence interval time is taken as the earliest release time t of the aircraft, and at the same time, let , , , go to step 3.2 to continue the iteration; Step 3.4: Output the aircraft integrated scheduling timing allocation plan. 8. The method for scheduling aircraft offshore platform operations according to claim 3, characterized in that: the step S4 adopts a binary tournament strategy to select the parent population, which is as follows: S41. Determine the number of individuals to be selected each time: In the binary tournament selection, two individuals are selected for comparison each time; S42, randomly select individuals: randomly select 50 individuals from the population initialized in step S2; S43, selecting individuals according to fitness values: according to the total time used by the scheduling scheme of each individual, selecting the individual with the shortest total time used by the scheduling scheme to enter the next generation population; S44, repeat the selection process: repeat the above steps to the maximum number of evaluations, until the size of the parent population reaches the size of the initialization population. 9. The method for scheduling aircraft offshore platform operations according to claim 3, characterized in that: the step S5 uses crossover and mutation operations to update the individuals in the parent population of step S4, specifically as follows: Crossover operation: For the three-level coding of transfer priority, maintenance and service support process, and departure, two-point crossover is performed by randomly selecting two intersection points; for the coding of parking space layout, single-point crossover is adopted. If there are duplicate genes after crossover, the duplicate genes are deleted and selected again from the available optional parking spaces; Mutation operation: For the three-level codes of transfer priority, maintenance and service support procedures, and dispatch and departure, random key mutation is performed with a random probability of 5%; for the parking bay layout code, in order to meet the extension update and internal optimization of the gene search domain, a 50% probability of performing random key mutation and a 50% probability of performing exchange mutation are set; Two strategies are used to update individual genes and expand the search range of solutions: Individual exchange strategy: synchronously select two aircraft within an individual to exchange gene fragments of the same process, select an individual gene to exchange the priority of the same process of different aircraft, and form a new individual gene; Individual recombination strategy: randomly select an aircraft within the individual to reorganize the gene fragments of the process, select an individual gene to reorganize the priorities of different processes of the same carrier-based aircraft, and form a new individual gene. 10. The method for scheduling aircraft offshore platform operations according to claim 3, characterized in that: the offspring individuals in step S7 calculate the comprehensive efficiency The details are as follows: , where: Cmax is the historical maximum value of the completion time, Cmin is the historical minimum value of the completion time, and C is the completion time of the offspring individual calculated this time; after calculating the comprehensive efficiency, the individual with the shortest total time used in the scheduling scheme is selected and compared with the parent individual with the shortest total time used in the scheduling scheme in step S4, and the historical best individual is retained; The specific steps of step S8 are as follows: after retaining the historical best individual in step S7, replace the individual with the lowest fitness in the population with the best individual to form a new individual coding population, repeat steps S5 to S7 for the new individual coding population until the number of decoding evaluations meets the iterative termination condition of the maximum number of evaluations, then end the decoding and output the scheduling plan corresponding to the historical best individual.