A MAS model for aviation equipment support and a method for analyzing equipment configuration requirements

Abstract

The present application relates to an aviation equipment support MAS model and a support equipment configuration demand analysis method, which belongs to the field of aviation support technology, including a support task agent, a support process agent, an aviation equipment agent, a support equipment agent, a random dynamic programming agent, and an effectiveness evaluation agent. Various intelligent agents are independent of each other and communicate and cooperate with each other to jointly complete the aviation equipment support task, and determine the optimal quantity combination plan of the support equipment through the evaluation and analysis of the task effectiveness. At the same time, due to the integration of random dynamic programming, this method can reduce the amount of calculation and improve the analysis efficiency.

Description

Technical Field

[0001] The present invention relates to the field of aviation equipment support technology, and in particular to an aviation equipment support MAS model and a support equipment configuration demand analysis method. Background Art

[0002] Support equipment is an important part of the aviation equipment support system. As the material basis of aviation equipment support, the rationality of its matching during the entire life cycle of aviation equipment directly affects the support efficiency of the main equipment and even the generation and performance of combat effectiveness. Support equipment participating in aviation equipment support tasks must comply with certain support logic. Under the requirements of the types of support equipment and the timing of use constrained by this logic, accurately determining the number of support equipment can effectively ensure the rationality of support equipment configuration.

[0003] Whether it is the planning and demonstration in the aviation equipment development stage or the in-service assessment and combat evaluation in the use stage, the demand analysis of aviation support equipment configuration has received increasing attention. However, due to the complexity and diversity of support mission types and the number of support equipment, most studies are still limited to qualitative analysis, quantitative analysis is rare, and Agent technology has never been involved. An agent is a computing entity that resides in a certain environment, can function continuously and autonomously, has characteristics such as residence, responsiveness, sociality, initiative, and can interact with any other machine. It not only has the ability to solve knowledge, but also can acquire knowledge in a dynamic and uncertain environment in a timely manner. It is very suitable for problems in the fields of performance evaluation, planning and scheduling. In view of the above, the present invention proposes an aviation equipment support MAS model and a support equipment configuration demand analysis method to improve the rationality of support equipment configuration. Summary of the invention

[0004] The present invention aims at the problem of calculating the configuration requirements of support equipment for aviation equipment with two-stage and multi-process conflicting support, and proposes an aviation equipment support MAS model and a support equipment configuration requirement analysis method. Support equipment participates in aviation equipment support tasks according to the support logic. For two-stage and multi-process conflicting support, the two stages refer to the existence of a sequential relationship between different support equipment, that is, only after one or several support equipment complete the support, the remaining support equipment can participate in the support task. Multi-process conflict means that in the same stage, the support equipment is in an uncertain order parallel relationship, but cannot be supported at the same time. MAS is a collection of multiple agents. In the present invention, six types of agents are established: support task agent, support process agent, aviation equipment agent, support equipment agent, random dynamic programming agent and performance evaluation agent. These six types of agents cooperate with each other to complete the support task together, and determine the optimal combination of the number of support equipment through the evaluation and analysis of task effectiveness.

[0005] In order to achieve the above purpose, the embodiment of the present invention provides the following technical solutions:

[0006] First,

[0007] An aviation equipment support MAS model, the MAS model comprising: a support task agent, a support process agent and a random dynamic programming agent;

[0008] The support task Agent is used to clarify the number of aviation equipment N that needs to be supported, as well as the types of support equipment involved in the support task. For each type of support equipment, it generates several Airamka distribution or exponential distribution random numbers of support operation time based on the elimination method, and sends the number of aviation equipment N and the types of support equipment to the support process Agent;

[0009] The support process agent is used to receive the number of aviation equipment N and the types of support equipment sent by the support task agent, clarify the support logic of the support equipment participating in the support task, that is, distinguish the use stage and use relationship of the support equipment, and send the types of support equipment participating in the support task in the first stage to the random dynamic programming agent;

[0010] The random dynamic programming agent is used to receive the types of support equipment participating in the first phase support mission sent by the support process agent, call the random number of related support operation time and the number of aviation equipment generated by the support task agent, determine the optimal quantity combination plan of the first phase support equipment through the constructed random dynamic programming model, and send the determined optimal quantity combination plan of the first phase support equipment to the support process agent;

[0011] The support process agent determines the quantity of each support equipment according to the final quantity combination plan of the first-stage support equipment received and the quantity range of the second-stage support equipment, and dispatches the support equipment to support the aviation equipment, and simultaneously records the status information of the support equipment and aviation equipment.

[0012] Furthermore, the MAS model also includes a support equipment agent and an aviation equipment agent;

[0013] The support equipment agent is used to receive the work instructions sent by the support process agent, randomly send a support request to any idle aviation equipment agent, and after the request is confirmed, call the random number of support operation time generated by the support task agent to support the confirmed aviation equipment agent;

[0014] The aviation equipment agent is used to receive the support request sent by the support equipment agent, randomly determine which type and which support equipment to provide support for it, send confirmation information to the support equipment agent, and record the start and end time of the operation, and synchronize the support operation status with the support process agent.

[0015] Furthermore, the MAS model also includes a performance evaluation agent. After the guarantee operation is completed, the guarantee process agent sends the start and end time of the operation to the performance evaluation agent. After the entire guarantee task is completed, the performance evaluation agent performs performance evaluation.

[0016] Second,

[0017] A method for analyzing equipment configuration requirements based on the aviation equipment support MAS model as described in the above aspects, the method comprising the following specific steps:

[0018] Step 1: The support task agent specifies the number of aviation equipment N that needs to be supported, as well as the types of support equipment involved in the support task; for each type of support equipment, a number of Airamka distribution or exponential distribution random numbers of support operation time are generated based on the elimination method;

[0019] Step 2: The support task agent sends the number of aviation equipment N and the types of support equipment to the support process agent; after receiving the support process agent, it clarifies the support logic of the support equipment participating in the support task, that is, distinguishes the use stage and use relationship of the support equipment, and divides the support equipment into the first-stage support equipment and the second-stage support equipment. If the first-stage support equipment is in an uncertain order and parallel use relationship, but cannot work at the same time, the support process agent sends the first-stage support equipment to the random dynamic programming agent;

[0020] Step 3: For the first-stage support equipment, the random dynamic programming agent calls the random number of related support operation time and the number of aviation equipment generated by the support task agent, and determines the optimal quantity combination plan of the first-stage support equipment through the random dynamic programming model.

[0021] Furthermore, the method further comprises the following specific steps:

[0022] Step 4: The random dynamic programming agent sends the determined optimal quantity combination plan of the first-stage support equipment to the support process agent, and the support process agent uses any optimal quantity combination plan to initialize the first-stage support equipment agent;

[0023] Step 5: Ensure that the process agent Initialize the second stage support equipment Agent, where They represent the number of r-item support equipment in the first phase respectively;

[0024] Step 6: According to the determined support logic, the support process agent sends a work order to the idle first-stage support equipment agent, and the first-stage support equipment agent randomly sends a support request to any idle aviation equipment agent;

[0025] Step 7: After receiving the support request, any aviation equipment agent randomly determines which type and which support equipment will provide support for it, sends a confirmation message to the support equipment agent, and records the start time of the operation;

[0026] Step 8: After receiving the confirmation information, the selected support equipment Agent calls the random number of the support operation time generated by the support task Agent as the current operation time, and provides support for the aviation equipment Agent;

[0027] Step 9: After the support operation is completed, the aviation equipment agent updates its own status to "the support operation is completed" and sends it to the support process agent synchronously;

[0028] Step 10: After the support process agent updates the status of the aviation equipment agent, it sends a work order to the remaining first-stage support equipment agents that are not supported, and repeats steps 6 to 9 until the aviation equipment agent completes all first-stage support operations;

[0029] Step 11: For the second-stage support equipment, after the first-stage support operation is completed, the support process agent sends the work order to the second-stage support equipment agent according to the determined support logic. The second-stage support equipment agent refers to the first-stage support equipment agent and repeats steps 6 to 10 to perform the second-stage support operations.

[0030] Furthermore, in step three, the method for constructing the stochastic dynamic programming model includes the following specific steps:

[0031] 1) Construct the objective function:

[0032]

[0033]

[0034] 2) Construct constraints:

[0035]

[0036]

[0037]

[0038]

[0039]

[0040] Where, M = (1, 2, ..., m) is the set of support equipment participating in the first phase, N is the number of aviation equipment, They represent the time required for m items of support equipment to complete operations for all aviation equipment in the first phase, They represent the number of m items of support equipment in the first phase, They represent the support operation time of different support equipment m items for aviation equipment in the first phase, is a decision variable, and its value is:

[0041]

[0042] 3) The expected value model is used to optimize the stochastic planning and obtain the optimal quantity combination plan for the first phase of support equipment.

[0043] Furthermore, the method further comprises the following specific steps:

[0044] Step 12: Repeat steps 6 to 11 for the two-stage support equipment quantity combinations determined in steps 4 and 5;

[0045] Step 13: For any optimal combination of equipment quantity for the first phase determined in step 3, Repeat steps 6 to 12 within the range to conduct sensitivity analysis on the number of equipment supported in the first phase;

[0046] Step 14: Repeat steps 6 to 13 to traverse all the optimal quantity combination plans for the first-phase support equipment determined in step 3.

[0047] Furthermore, the method further comprises the following specific steps:

[0048] Step 15: After the traversal is completed, the guarantee process agent sends a request message to the performance evaluation agent, and the performance evaluation agent performs performance evaluation;

[0049] Step 16: Compare the mission time, average queuing waiting time and mission time ratio of different combinations of support equipment quantities, and comprehensively evaluate and select the optimal combination of support equipment quantities.

[0050] Furthermore, the performance evaluation agent performs performance evaluation, including the following specific steps:

[0051] 1) For all aviation equipment agents, compare the start time of each support operation, take the minimum value as the task start time; compare the end time of each support operation, take the maximum value as the task end time, and subtract the two to get the task time;

[0052] 2) For any aviation equipment Agent, the difference between the start time of the next support operation and the end time of the previous support operation is the queue waiting time;

[0053] 3) Accumulate the waiting time of all aviation equipment Agents in queue and compare it with the number of aviation equipment N to get the average waiting time;

[0054] 4) Perform normal distribution / lognormal distribution test and parameter estimation on task time and average queue waiting time, and take 0.95 quantile as the characteristic value of task time and average queue waiting time;

[0055] 5) Estimate the parameters of the Cauchy distribution for the ratio of the average waiting time in a queue to the task time, and take the 0.95 quantile as the characteristic value of the ratio of the average waiting time in a queue to the task time.

[0056] Beneficial effects of the invention:

[0057] Aiming at the two-stage multi-process conflicting aviation equipment support, the MAS model is used to evaluate the support efficiency of the support equipment, and then the optimal quantity combination plan of the support equipment is given to realize the quantitative analysis of the support equipment configuration requirements; at the same time, the optimal quantity combination of the support equipment in the first stage is analyzed based on the random dynamic programming method, which can reduce the amount of calculation and improve the analysis efficiency.

[0058] These and other objects, features and advantages of the present invention will be fully reflected in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 The schematic diagram of the framework of the MAS model for aviation equipment support in this application is shown.

[0060] Figure 2 A schematic diagram of random numbers of the Airamka distribution for the equipment configuration requirements analysis method of the present application is shown.

[0061] Figure 3 The Gantt chart of the launch support for the equipment configuration demand analysis method of this application is shown.

[0062] Figure 4 The histogram of the launch support time and average time of the support equipment configuration demand analysis method of this application is shown.

[0063] Figure 5A schematic diagram showing the trend of the release support time changing with the number of bomb-hanging vehicles in the support equipment configuration demand analysis method of this application is shown.

[0064] Figure 6 A schematic diagram showing the trend of average waiting time changing with the number of bomb-hanging vehicles in the equipment configuration demand analysis method of this application is shown.

[0065] Figure 7 A schematic diagram showing the trend of the effective support time ratio changing with the number of bomb-hanging vehicles in the support equipment configuration demand analysis method of this application. DETAILED DESCRIPTION

[0066] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.

[0067] It should be understood by those skilled in the art that, in the disclosure of the specification, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. Therefore, the above terms should not be understood as limiting the present invention.

[0068] It is to be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element may be one, while in another embodiment, the number of the element may be multiple, and the term "one" should not be understood as a limitation on the quantity.

[0069] Embodiment 1,

[0070] Combination Figure 1 , an aviation equipment support MAS model, the MAS model includes: a support task agent, a support process agent, a random dynamic programming agent, a support equipment agent, an aviation equipment agent and an effectiveness evaluation agent.

[0071] The support task Agent is used to clarify the number of aviation equipment N that needs to be supported, as well as the types of support equipment involved in the support task. For each type of support equipment, it generates several Airamka distribution or exponential distribution random numbers of support operation time based on the elimination method, and sends the number of aviation equipment N and the types of support equipment to the support process Agent;

[0072] The support process Agent is used to receive the number N of aviation equipment and the types of support equipment sent by the support task Agent, clarify the support logic of the support equipment participating in the support task, that is, distinguish the use stage and use relationship of the support equipment, and send the types of support equipment participating in the support task in the first stage to the random dynamic programming Agent.

[0073] The random dynamic programming agent is used to receive the types of support equipment participating in the first phase support mission sent by the support process agent, call the random number of related support operation time and the number of aviation equipment generated by the support task agent, determine the optimal quantity combination plan of the first phase support equipment through the constructed random dynamic programming model, and send the determined optimal quantity combination plan of the first phase support equipment to the support process agent.

[0074] The support process agent determines the quantity of each support equipment according to the final quantity combination plan of the first-stage support equipment received and the quantity range of the second-stage support equipment, and dispatches the support equipment to support the aviation equipment, and simultaneously records the status information of the support equipment and aviation equipment.

[0075] The support equipment agent is used to receive the work instructions sent by the support process agent, randomly send a support request to any idle aviation equipment agent, and after the request is confirmed, call the random number of support operation time generated by the support task agent to support the confirmed aviation equipment agent;

[0076] The aviation equipment agent is used to receive the support request sent by the support equipment agent, randomly determine which type and which support equipment to provide support for it, send confirmation information to the support equipment agent, and record the start and end time of the operation, and synchronize the support operation status with the support process agent.

[0077] After the guarantee operation is completed, the guarantee process agent sends the start and end time of the operation to the performance evaluation agent. After the entire guarantee task is completed, the performance evaluation agent performs performance evaluation.

[0078] The support task agent specifies the number of aviation equipment N that needs to be supported, as well as the types of support equipment participating in the support task. For each type of support equipment, several random numbers of the support operation time are generated based on the elimination method using the Airamka distribution or exponential distribution. The support task agent sends the number of aviation equipment N and the types of support equipment to the support process agent. After receiving the support process agent, the support logic of the support equipment participating in the support task is clarified, that is, the use phase and use relationship of the support equipment are distinguished. If the support equipment in the first stage is in an uncertain order and parallel, but the use relationship cannot work at the same time, the support process agent sends the support equipment in the first stage to the random dynamic programming agent. For the support equipment in the first stage, the random dynamic programming agent calls the random numbers of the relevant support operation time and the number of aviation equipment generated by the support task agent, and determines the optimal quantity combination plan for the first stage support equipment through the random dynamic programming model.

[0079] The random dynamic programming agent sends the optimal quantity combination scheme of the first-stage support equipment to the support process agent, and the support process agent uses any optimal quantity combination scheme to initialize the first-stage support equipment agent. Initialize the second stage support equipment agent, where Respectively represent the number of r items of support equipment in the first stage. According to the determined support logic, the support process agent sends a work instruction to the idle first-stage support equipment agent, and the first-stage support equipment agent randomly sends a support request to any idle aviation equipment agent. If the first-stage support equipment is in an uncertain order parallel relationship, but cannot be used at the same time, any aviation equipment agent will randomly determine which type and which support equipment to provide support for it after receiving the support request, and send a confirmation message to the support equipment, and record the start time of the operation. After receiving the confirmation message, the selected support equipment agent calls the random number of the support operation time generated by the support task agent as the operation time of this operation, and provides support for the aviation equipment agent. After the support operation is completed, the aviation equipment agent records the end time of the operation, and sends the start and end time of the operation to the performance evaluation agent. At the same time, the aviation equipment agent updates its own status as the completion of the support operation, and sends it to the support process agent synchronously.

[0080] The support task agent, support process agent, aviation equipment agent, support equipment agent, random dynamic programming agent and effectiveness evaluation agent work together to complete the support task and determine the optimal combination of the number of support equipment. Aiming at the actual two-stage multi-process conflict support of aviation equipment, based on random dynamic programming and MAS technology, an effectiveness evaluation method for aviation support equipment configuration is created, forming a complete technical route for aviation support equipment configuration measurement, which can be used for quantitative analysis of its needs and provide theoretical support for the agile support of aviation equipment.

[0081] Embodiment 2,

[0082] refer to Figures 2 to 7 According to a preferred embodiment of the present invention, a support equipment configuration demand analysis method will be described in detail below. The optimization method includes the following specific steps: Step 1: The support task Agent specifies the number of aviation equipment N that needs to be supported and the types of support equipment participating in the support task. For each type of support equipment, a number of Airamka distribution or exponential distribution random numbers of support operation time are generated based on the elimination method.

[0083] Step 2: The support task agent sends the number of aviation equipment N and the types of support equipment to the support process agent. After receiving the support process agent, it clarifies the support logic of the support equipment participating in the support task, that is, distinguishes the use stage and use relationship of the support equipment. If the first stage support equipment is in an uncertain order parallel but cannot work at the same time, the support process agent sends the first stage support equipment to the random dynamic programming agent.

[0084] Step 3: For the first phase support equipment, the random dynamic programming agent calls the random number of support operation time and the number of aviation equipment generated by the support task agent, and determines the optimal number combination plan of the first phase support equipment through the random dynamic programming model. The construction of the random dynamic programming model includes the following specific steps:

[0085] 1) Construct the objective function:

[0086]

[0087]

[0088] 2) Construct constraints:

[0089]

[0090]

[0091]

[0092]

[0093]

[0094] Where, M = (1, 2, ..., m) is the set of support equipment participating in the first phase, N is the number of aviation equipment, They represent the time required for m items of support equipment to complete operations for all aviation equipment in the first phase, They represent the number of m items of support equipment in the first phase, They represent the support operation time of different support equipment m items for aviation equipment in the first phase, is a decision variable, and its value is:

[0095]

[0096] 3) The expected value model is used to optimize the stochastic planning and obtain the optimal quantity combination plan for the first phase of support equipment.

[0097] Step 4: The random dynamic programming agent sends the determined optimal quantity combination plan of the first-stage support equipment to the support process agent, and the support process agent uses any optimal quantity combination plan to initialize the first-stage support equipment agent.

[0098] Step 5: Ensure that the process agent Initialize the second stage support equipment Agent, where They respectively represent the number of r-item support equipment in the first phase.

[0099] Step 6: According to the determined support logic, the support process agent sends a work order to the idle first-stage support equipment agent, and the first-stage support equipment agent randomly sends a support request to any idle aviation equipment agent.

[0100] Step 7: After receiving the support request, any aviation equipment Agent randomly determines which type and which support equipment will provide support for it, sends a confirmation message to the support equipment Agent, and records the start time of the operation.

[0101] Step 8: After receiving the confirmation information, the selected support equipment Agent calls the random number of the support operation time generated by the support task Agent as the current operation time to provide support for the aviation equipment Agent.

[0102] Step 9: After the support operation is completed, the aviation equipment agent records the end time of the operation and sends the start and end time of the operation to the performance evaluation agent. At the same time, the aviation equipment agent updates its own status as the support operation has been completed and sends it to the support process agent simultaneously.

[0103] Step 10: After the support process agent updates the status of the aviation equipment agent, it sends work instructions to the remaining first-stage support equipment agents that are not supported, and repeats steps 6 to 9 until the aviation equipment agent completes all first-stage support operations.

[0104] Step 11: After the first phase of support operations are completed, the support process agent sends the work instructions to the second phase support equipment agent according to the determined support logic. The second phase support equipment agent refers to the first phase support equipment agent and repeats steps 6 to 10 to perform the second phase of support operations.

[0105] Step 12: Repeat steps 6 to 11 100 times for the two-stage support equipment quantity combination determined in steps 4 and 5.

[0106] Step 13: For any optimal combination of equipment quantity for the first phase determined in step 3, Repeat steps 6 to 12 within the range to conduct sensitivity analysis on the number of support equipment in the first phase.

[0107] Step 14: Repeat steps 6 to 13 to traverse all the optimal quantity combination plans for the first-phase support equipment determined in step 3.

[0108] Step 15: After the traversal is completed, the guarantee process agent sends a request message to the performance evaluation agent, and the performance evaluation agent performs performance evaluation.

[0109] Step 16: Compare the mission time, average queuing waiting time and mission time ratio of different combinations of support equipment quantity, and comprehensively evaluate and select the optimal combination of support equipment quantity. The performance evaluation agent performs performance evaluation, including the following specific steps:

[0110] 1) For all aviation equipment agents, compare the start time of each support operation, take the minimum value as the task start time; compare the end time of each support operation, take the maximum value as the task end time, and subtract the two to get the task time;

[0111] 2) For any aviation equipment Agent, the difference between the start time of the next support operation and the end time of the previous support operation is the queue waiting time;

[0112] 3) Accumulate the waiting time of all aviation equipment Agents in queue and compare it with the number of aviation equipment N to get the average waiting time;

[0113] 4) Perform normal distribution / lognormal distribution test and parameter estimation on task time and average queue waiting time, and take 0.95 quantile as the characteristic value of task time and average queue waiting time;

[0114] 5) Estimate the parameters of the Cauchy distribution for the ratio of the average waiting time in a queue to the task time, and take the 0.95 quantile as the characteristic value of the ratio of the average waiting time in a queue to the task time.

[0115] Follow steps 1 to 16 and take aviation equipment launch support as an example to verify this method.

[0116] Flight support is a support task implemented before the aviation equipment is dispatched. This stage mainly completes external inspections and "charging, filling, adding, and hanging" work. Among them, the support equipment required for "charging, filling, adding, and hanging" are power supply vehicles, oxygen / nitrogen / air conditioning vehicles, refueling vehicles, and bomb hanging vehicles. The support logic is: oxygen / nitrogen / air conditioning, power supply, and refueling are not in order, but cannot be carried out at the same time. Bomb hanging can only be carried out after the first three support items are completed. That is, "charging, filling, and adding" can be regarded as support stage I, and "hanging" is support stage II.

[0117] Assume that 10 aircraft of a certain type are launched for support. The power supply vehicle, refueling vehicle, bomb loading vehicle, and oxygen / nitrogen / air conditioning vehicle involved in the support are located 50m, 40m, 30m, and 20m behind the aircraft respectively. The vehicle moving speed in the field is 10km / h. After the support begins, the vehicle randomly selects an aircraft for support. During the simulation experiment, the time unit is s.

[0118] Assume that the operation time of "charging, filling, adding, and hanging" follows the Elam Ka distribution. Based on the rejection method, Elam Ka distribution random numbers with θ = 600 in the interval [480,720], θ = 900 in the interval [720,1080], and θ = 180 in the interval [120,240] are generated as the operation time of power-on and refueling, hanging bombs, and oxygen / nitrogen / cooling in the simulation experiment, in seconds, such as Figure 2 As shown, Figure 2 The first picture shows the support operation time of the power supply vehicle, the second picture shows the support operation time of the oxygen / nitrogen / air conditioning vehicle, the third picture shows the support operation time of the refueling vehicle, and the fourth picture shows the support operation time of the bomb hanging vehicle.

[0119] Substituting the above random numbers into the random dynamic programming model, it is found through optimization calculation that the objective function can be optimized, that is, the shortest operation time is guaranteed in phase I;

[0120] The results of random dynamic programming optimization are shown in Table 1:

[0121] Table 1

[0122]

[0123] Based on the four configurations of power supply vehicles, oxygen / nitrogen / air conditioning vehicles, and refueling vehicles shown in Table 1, the number of bomb-hanging vehicles is taken in the range of [2,9] with a step size of 1. The release support time and average queue waiting time are used as output indicators to analyze the impact of different configurations on the support efficiency. The random service principle of the simulation process and the randomness of the "charging, filling, adding, and hanging" operation time determine the non-uniqueness of the calculation results. For this reason, the simulation is run 100 times under different configurations. Taking the configurations of power supply vehicles, oxygen / nitrogen / air conditioning vehicles, and refueling vehicles as 4, 1, and 5 as an example, when the number of bomb-hanging vehicles is 2, the release support Gantt chart can be obtained after one run as follows Figure 3 The results of running 100 times are shown in Figure 4 shown.

[0124] Combination Figure 4 , when the "charging, filling, adding, and hanging" support operation time all obey the Airamka distribution, the outer contour curves of the launch support time and the average waiting time are close to the normal distribution or log-normal distribution. To this end, the simulation results of the launch support time and the average waiting time under the support equipment configurations of different power supply vehicles, nitrogen / oxygen / air conditioning vehicles, and refueling vehicles, and the bomb hanging vehicles are 2-9 in sequence as shown in Table 1 are respectively tested for distribution.

[0125] Based on the least squares estimation, the normal distribution fitting test is performed on the release guarantee time under different configurations. The test results are shown in Table 2. It can be seen that under different numbers of vehicles, the fitting results have a good linear trend, that is, the release guarantee time obeys the normal distribution. The release guarantee time under the 0.95 quantile is listed in Table 2, and its change trend is as follows Figure 5 shown.

[0126] The normal distribution fitting test results of the launch support time under different configurations are shown in Table 2:

[0127] Table 2

[0128]

[0129]

[0130]

[0131]

[0132] Combination Figure 5,Under the four different configurations of power supply vehicles, nitrogen / oxygen / air conditioning vehicles, and refueling vehicles, the release support time is not much different, and it shows basically the same trend with the change in the number of bomb hanging vehicles, that is, with the increase in the number, the release support time first decreases rapidly, and then gradually stabilizes, especially after reaching 5 vehicles, the support time does not change much. From this perspective, the increase in vehicles will not bring about a significant improvement in support efficiency. It can be concluded that under different numbers of power supply vehicles, nitrogen / oxygen / air conditioning vehicles, and refueling vehicles, there is obviously an optimal solution for the configuration of bomb hanging vehicles.

[0133] Obviously, the working logic of the release support will cause idle waiting of various support equipment. For this reason, the average waiting time is selected as a comprehensive measure of the waiting time of the four types of support equipment. Also based on the least squares estimation, the average waiting time of various support equipment is tested for distribution when the number of the four power supply vehicles, nitrogen / oxygen / air conditioning vehicles, and refueling vehicles shown in Table 1 is 2-9, and the bomb hanging vehicles are configured respectively. The test results are shown in Table 3. It can be seen that under different numbers of vehicles, the fitting results have a good linear trend, that is, the average queuing time also obeys the normal distribution well. The average queuing time at the 0.95 quantile is listed in Table 3, and its changing trend is as follows Figure 6 shown.

[0134] The normal distribution fitting test results of the average waiting time under different configurations are shown in Table 3:

[0135] Table 3

[0136]

[0137]

[0138]

[0139] Combination Figure 6 Under a certain configuration of power supply truck, nitrogen / oxygen / air conditioning truck, and refueling truck, the average waiting time generally shows an upward trend of first slow and then fast as the number of bomb-hanging trucks increases. In comparison, when the number of power supply trucks, nitrogen / oxygen / air conditioning trucks, and refueling trucks is 5, 1, 4 or 4, 1, 5, the average waiting time is generally longer. The reason is that although the power-on and refueling operation time is longer than the inflation operation time, the nitrogen / oxygen / air conditioning trucks are too few, which prolongs the time for aviation equipment to wait for inflation support; when the number of power supply trucks, nitrogen / oxygen / air conditioning trucks, and refueling trucks is 4, 2, and 4, respectively, the inflation operation is more flexible due to the slight reduction in the number of power supply trucks or refueling trucks with long operation time and the increase in the number of nitrogen / oxygen / air conditioning trucks, and the average waiting time is reduced; when the number of power supply trucks, nitrogen / oxygen / air conditioning trucks, and refueling trucks is 4, 1, and 4, respectively, on the basis of the comparative relationship between the power-on, refueling, and inflation operation time, the redundant configuration is maximized and the shortest average waiting time is achieved.

[0140] In summary, the average waiting time varies among the four configurations of power supply vehicle, nitrogen / oxygen / air conditioning vehicle, and refueling vehicle. However, considering that the release guarantee time is also different, the average waiting time cannot be directly used as the basis for judging which configuration is better. A more appropriate method should be to judge the proportion of effective guarantee time in the entire release guarantee process, that is, 1-t 等待 / T 放飞 Obviously, the higher the effective guarantee time ratio is, the higher the guarantee efficiency is. Since the average waiting time and the release guarantee time are both subject to normal distribution, their ratio t 等待 / T 放飞 It obeys the Cauchy distribution. Therefore, based on the same idea, on the basis of parameter estimation of the Cauchy distribution, the proportion of effective protection time is calculated using the 0.95 quantile.

[0141] As a continuous distribution with no expectation and variance, the parameter estimation of the Cauchy distribution, classical estimation methods such as moment estimation method and maximum likelihood estimation method are not advisable, but from its cumulative distribution function, we can see that:

[0142]

[0143] Where: μ is the location parameter, λ is the scale parameter. Thus, we can get:

[0144]

[0145] The quantile estimation method can be used to 等待 / T 放飞 Parameter estimation of Cauchy distribution. The parameter estimation results and 0.95 quantiles are listed in Table 4 for the configurations of four power supply vehicles, nitrogen / oxygen / air conditioning vehicles, and refueling vehicles, and the number of bomb hanging vehicles is 2-9. The trend of the effective support time ratio under each configuration with the number of bomb hanging vehicles is shown in Table 4. Figure 7 .

[0146] The estimated results of the Cauchy distribution parameters of the ratio of the average waiting time to the release guarantee time under different configurations are shown in Table 4:

[0147] Table 4

[0148]

[0149]

[0150]

[0151] Figure 7 and Figure 6Different trends are shown. When the number of power trucks, nitrogen / oxygen / air conditioning trucks, and refueling trucks is 4, 1, and 4, the average waiting time is relatively short and the effective guarantee time accounts for a relatively high proportion. Obviously, the configuration is better. When the number of power trucks, nitrogen / oxygen / air conditioning trucks, and refueling trucks is 5, 1, and 4, although the average waiting time is relatively the highest, the effective guarantee time accounts for a suboptimal level. The average waiting time is at an intermediate level for power trucks, nitrogen / oxygen / air conditioning trucks, and refueling trucks with 4, 2, and 4 and 4, 1, and 5, respectively. The effective guarantee time accounts for a relatively poor proportion and there is no obvious difference. Figure 5 The change in the launch support time under the four configurations is not significant. Obviously, 4 power supply vehicles, 1 nitrogen / oxygen / air conditioning vehicle, and 4 refueling vehicles are the best configuration. The increase in the number of bomb hanging vehicles will reduce the launch support time on the one hand, but on the other hand, the support process at the end of the support operation also leads to an increase in the average waiting time, and the resulting decrease in the proportion of effective support time. Figure 5 and Figure 7 According to the changing trend, when there are 6 bomb-carrying vehicles, it can ensure a shorter release guarantee time and the proportion of effective guarantee time is also in the middle state, which can be the first choice for its number.

[0152] This invention is a two-stage multi-process conflict assurance for aviation equipment. Based on random dynamic programming and MAS technology based on the principle of random service, a method for analyzing the configuration requirements of assurance equipment is proposed. The flight assurance is used as a verification. Through random number generation, optimization calculation, Monte Carlo simulation and distribution test, the influence of assurance equipment configuration on assurance effectiveness is analyzed. The results show that this method can reflect the emergence law of assurance equipment subsystems, can be used for quantitative analysis of assurance equipment requirements, and can provide theoretical support for the agile assurance of aviation equipment.

[0153] It should be understood by those skilled in the art that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The advantages of the present invention have been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments, and the embodiments of the present invention may be deformed or modified in any way without departing from the principles.

Claims

1. An aviation equipment support MAS model system, characterized in that: The MAS model system includes: a security task agent, a security process agent and a random dynamic programming agent; The support task Agent is used to clarify the number of aviation equipment N that needs to be supported, as well as the types of support equipment involved in the support task. For each type of support equipment, it generates several Airamka distribution or exponential distribution random numbers of support operation time based on the elimination method, and sends the number of aviation equipment N and the types of support equipment to the support process Agent; The support process agent is used to receive the number of aviation equipment N and the types of support equipment sent by the support task agent, clarify the support logic of the support equipment participating in the support task, that is, distinguish the use stage and use relationship of the support equipment, and send the types of support equipment participating in the support task in the first stage to the random dynamic programming agent; The random dynamic programming agent is used to receive the types of support equipment participating in the first phase support mission sent by the support process agent, call the random number of related support operation time and the number of aviation equipment generated by the support task agent, determine the optimal quantity combination plan of the first phase support equipment through the constructed random dynamic programming model, and send the determined optimal quantity combination plan of the first phase support equipment to the support process agent; The support process agent determines the quantity of each support equipment according to the received optimal quantity combination plan of the first-stage support equipment and the quantity range of the second-stage support equipment, dispatches the support equipment to support the aviation equipment, and simultaneously records the status information of the support equipment and aviation equipment.

2. The aviation equipment support MAS model system according to claim 1, characterized in that: The MAS model system also includes a support equipment agent and an aviation equipment agent; The support equipment agent is used to receive the work instructions sent by the support process agent, randomly send a support request to any idle aviation equipment agent, and after the request is confirmed, call the random number of support operation time generated by the support task agent to support the confirmed aviation equipment agent; The aviation equipment agent is used to receive the support request sent by the support equipment agent, randomly determine which type and which support equipment to provide support for it, send confirmation information to the support equipment agent, and record the start and end time of the operation, and synchronize the support operation status with the support process agent.

3. The aviation equipment support MAS model system according to claim 2, characterized in that: The MAS model system also includes a performance evaluation agent. After the guarantee operation is completed, the guarantee process agent sends the start and end time of the operation to the performance evaluation agent. After the entire guarantee task is completed, the performance evaluation agent performs performance evaluation.

4. A method for analyzing equipment configuration requirements applied to the aviation equipment support MAS model system according to any one of claims 1 to 3, characterized in that: The method comprises the following specific steps: Step 1: The support task Agent specifies the number of aviation equipment N that needs to be supported, as well as the types of support equipment involved in the support task; for each type of support equipment, a number of Airamka distribution or exponential distribution random numbers of support operation time are generated based on the elimination method; Step 2: The support task agent sends the number of aviation equipment N and the types of support equipment to the support process agent; after receiving the support process agent, it clarifies the support logic of the support equipment participating in the support task, that is, distinguishes the use stage and use relationship of the support equipment, and divides the support equipment into the first-stage support equipment and the second-stage support equipment. If the first-stage support equipment is in an uncertain order and parallel use relationship, but cannot work at the same time, the support process agent sends the first-stage support equipment to the random dynamic programming agent; Step 3: For the first-stage support equipment, the random dynamic programming agent calls the random number of related support operation time and the number of aviation equipment generated by the support task agent, and determines the optimal quantity combination plan of the first-stage support equipment through the random dynamic programming model.

5. The support equipment configuration demand analysis method according to claim 4, characterized in that: The method further comprises the following specific steps: Step 4: The random dynamic programming agent sends the determined optimal quantity combination plan of the first-stage support equipment to the support process agent, and the support process agent uses any optimal quantity combination plan to initialize the first-stage support equipment agent; Step 5: Ensure that the process agent Initialize the second stage support equipment Agent, where They represent the number of r-item support equipment in the first phase respectively; Step 6: According to the determined support logic, the support process agent sends a work order to the idle first-stage support equipment agent, and the first-stage support equipment agent randomly sends a support request to any idle aviation equipment agent; Step 7: After receiving the support request, any aviation equipment agent randomly determines which type and which support equipment will provide support for it, sends a confirmation message to the support equipment agent, and records the start time of the operation; Step 8: After receiving the confirmation information, the selected support equipment Agent calls the random number of the support operation time generated by the support task Agent as the current operation time, and provides support for the aviation equipment Agent; Step 9: After the support operation is completed, the aviation equipment agent updates its own status to "the support operation is completed" and sends it to the support process agent synchronously; Step 10: After the support process agent updates the status of the aviation equipment agent, it sends a work order to the remaining first-stage support equipment agents that are not supported, and repeats steps 6 to 9 until the aviation equipment agent completes all first-stage support operations; Step 11: For the second-stage support equipment, after the first-stage support operation is completed, the support process agent sends the work order to the second-stage support equipment agent according to the determined support logic. The second-stage support equipment agent refers to the first-stage support equipment agent and repeats steps 6 to 10 to perform the second-stage support operations.

6. The support equipment configuration demand analysis method according to claim 4, characterized in that: In step three, the method for constructing the stochastic dynamic programming model includes the following specific steps: 1) Construct the objective function: 2) Construct constraints: Where, M = (1, 2, ..., m) is the set of support equipment participating in the first phase, N is the number of aviation equipment, They represent the time required for m items of support equipment to complete operations for all aviation equipment in the first phase, They represent the number of m items of support equipment in the first phase, They represent the support operation time of different support equipment m items for aviation equipment in the first phase, is a decision variable, and its value is: 3) The expected value model is used to perform optimization calculations on the stochastic dynamic programming model to obtain the optimal quantity combination plan for the first phase of support equipment.

7. The support equipment configuration demand analysis method according to claim 6, characterized in that: The method further comprises the following specific steps: Step 12: Repeat steps 6 to 11 for the two-stage support equipment quantity combinations determined in steps 4 and 5; Step 13: For any optimal combination of equipment quantity for the first phase determined in step 3, Repeat steps 6 to 12 within the range to conduct sensitivity analysis on the number of equipment for support in the second phase; Step 14: Repeat steps 6 to 13 to traverse all the optimal quantity combination plans for the first-phase support equipment determined in step 3.

8. The support equipment configuration demand analysis method according to claim 7, characterized in that: The method further comprises the following specific steps: Step 15: After the traversal is completed, the guarantee process agent sends a request message to the performance evaluation agent, and the performance evaluation agent performs performance evaluation; Step 16: Compare the mission time, average waiting time in queue and mission time ratio of different combinations of support equipment quantities, and comprehensively evaluate and select the optimal combination of support equipment quantities.

9. The support equipment configuration demand analysis method according to claim 8, characterized in that: The performance evaluation agent performs performance evaluation, including the following specific steps: 1) For all aviation equipment agents, compare the start time of each support operation, take the minimum value as the task start time; compare the end time of each support operation, take the maximum value as the task end time, and subtract the two to get the task time; 2) For any aviation equipment Agent, the difference between the start time of the next support operation and the end time of the previous support operation is the queue waiting time; 3) Accumulate the waiting time of all aviation equipment Agents in queue and compare it with the number of aviation equipment N to get the average waiting time; 4) Perform normal distribution / lognormal distribution test and parameter estimation on task time and average queue waiting time, and take 0.95 quantile as the characteristic value of task time and average queue waiting time; 5) Estimate the parameters of the Cauchy distribution for the ratio of the average waiting time in a queue to the task time, and take the 0.95 quantile as the characteristic value of the ratio of the average waiting time in a queue to the task time.

Images