Aviation Emergency Rescue Real-time Task Evaluation System

Abstract

The present invention relates to a real-time task evaluation system for aviation emergency rescue, including: a rescue personnel behavior prediction module based on data mining, an aviation medical transportation task simulation module, and a real-time task evaluation module based on artificial intelligence. The rescue personnel behavior prediction module mainly uses big data technology to analyze and predict the rescue behaviors of rescue personnel participating in medical transportation. The task simulation module takes the prediction results as input, fully considers the professional skill levels and non-skill levels of rescue personnel, as well as various objective uncontrollable factors, conducts modeling and simulation, and generates simulation results. The real-time task intelligent evaluation module takes the simulation results as input, uses artificial intelligence technology to score and conduct intelligent evaluation, while predicting the rescue results, identifies possible emergencies and risks during the rescue process, and provides guidance for the execution of rescue tasks.

Description

Technical Field

[0001] The invention relates to rescue technology, and in particular to an aviation emergency rescue real-time task evaluation system. Background Art

[0002] With regard to rescue technology, there is currently a lack of an evaluation system that can effectively and comprehensively evaluate rescue efficiency. Summary of the invention

[0003] The purpose of the present invention is to provide an aviation emergency rescue real-time mission assessment system to solve the above-mentioned problems of the prior art.

[0004] The present invention provides an aviation emergency rescue real-time task evaluation system, which includes: a professional rescue team behavior prediction module, which is used to call the historical rescue data and historical training data of the professional rescue team, predict the behavior performance of the professional rescue team in this rescue mission, and generate professional rescue team behavior prediction result data, which is transmitted to the medical transport task simulation module; an aviation emergency medical transport task simulation module, which is used to integrate real-time rescue mission data, patient data, mission execution environment data, the professional rescue team behavior prediction result data, and combine the emergency situation generation and processing mechanism, coordination mechanism, and information transmission mechanism within the aviation emergency medical transport task simulation module to run the medical transport task simulation model to obtain the system internal human resources. The micro-state information of each state node of the medical transfer personnel and the macro-state information of the medical transfer action during the medical transfer process are collected, and the micro-state information and the macro-state information are transmitted to the real-time task evaluation module; the real-time task evaluation module calls the micro-state information and the macro-state information generated by the aviation emergency medical transfer task simulation module, adopts the expert intelligent scoring model based on data mining and the established evaluation mechanism, generates the rescue task execution effect evaluation result, and provides decision support for the responsible personnel in the system; at the same time, the micro-state information, the macro-state information and the rescue task execution effect evaluation result are integrated, and the risk warning model is used to identify the possible risks in the task execution process, and provide countermeasures.

[0005] The present invention provides support for the decision-making of the relevant person in charge of the hospital to accept the transfer task for reference. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 The module diagram of the aviation emergency rescue real-time task evaluation system of the present invention is shown;

[0007] Figure 2 Shown is a schematic diagram of a professional rescue team behavior prediction module;

[0008] Figure 3 It is a simulation module for aviation emergency medical rescue and transfer mission;

[0009] Figure 4 Shown is a schematic diagram of the aeromedical transport real-time mission assessment module. DETAILED DESCRIPTION

[0010] In order to make the purpose, content, and advantages of the present invention more clear, the specific implementation methods of the present invention are further described in detail below in conjunction with the accompanying drawings and examples.

[0011] Figure 1 The module diagram of the aviation emergency rescue real-time task evaluation system of the present invention is shown as follows: Figure 1 As shown, the aviation emergency rescue real-time task evaluation system of the present invention includes: a professional rescue team behavior prediction module: used to call the historical rescue data and historical training data of the professional rescue team, predict the behavior performance of the professional rescue team in this rescue mission, and generate professional rescue team behavior prediction result data, and transmit it to the medical transport task simulation module. The aviation emergency medical transport task simulation module: used to integrate real-time rescue mission data, patient data, mission execution environment data, the professional rescue team behavior prediction result data, and combine the emergency situation generation and processing mechanism, coordination mechanism, and information transmission mechanism inside the aviation emergency medical transport task simulation module to run the medical transport task simulation model, obtain the micro-state information of each state node of the system internal personnel during the medical transport process and the macro-state information of the medical transport action, and transmit the micro-state information and the macro-state information to the real-time task evaluation module. Real-time task evaluation module: calls the micro-state information and the macro-state information generated by the aviation emergency medical transport task simulation module, adopts the expert intelligent scoring model based on data mining and the established evaluation mechanism, generates the rescue task execution effect evaluation result, and provides decision support for the responsible personnel in the system; at the same time, integrates the micro-state information, the macro-state information and the rescue task execution effect evaluation result, uses the risk warning model to identify the possible risks in the task execution process, and provides countermeasures.

[0012] like Figure 1 As shown, the internal personnel of the system include: personnel of the transfer receiving hospital, personnel of the transfer requesting hospital and patients; the personnel of the transfer receiving hospital include the person in charge of the transfer receiving hospital, ground support forces and professional rescue teams: the ground support forces include security guards and ground maintenance personnel of the transfer receiving hospital; the professional rescue team includes pilots, random maintenance personnel, random doctors and random nurses; the personnel of the transfer requesting hospital include the person in charge of the transfer requesting hospital, security guards and medical staff of the transfer requesting hospital.

[0013] like Figure 1As shown, the historical training data includes skill training data and non-skill training data. The skill training data includes general skill data, professional skill data, and process proficiency training data; the non-skill training data includes communication and coordination ability, command ability, psychological stress, and emergency response ability training data.

[0014] like Figure 1 As shown, the aviation emergency medical transport task simulation module uses multi-agent simulation technology to abstract into a transport request hospital subsystem, a transport receiving hospital subsystem, a transport demand subsystem and an environment subsystem; the transport request hospital subsystem includes a transport request hospital building entity, a transport request hospital lawn entity, a transport request hospital person in charge Agent, a transport request hospital security Agent, a doctor Agent, and a nurse Agent; the transport receiving hospital subsystem includes a transport receiving hospital building entity, a transport request hospital lawn entity, a stretcher object, an airborne equipment object, a take-off battery object, a transport request hospital person in charge Agent, a transport request hospital security Agent, a random doctor Agent, a random doctor Agent, a random nurse Agent, a helicopter Agent, a ground maintenance Agent, a random maintenance Agent, and a pilot Agent; the transport demand subsystem includes a patient Agent; the environment subsystem includes an accident generation Agent, a man-made accident Agent, and an environment control Agent.

[0015] like Figure 1 As shown, the aviation emergency medical transport task simulation module divides the medical transport process into the following state nodes: transport application, linkage mechanism activation, rescue helicopter rushing to the transport point, patient boarding, air medical treatment, patient disembarking, and medical treatment in the hospital area;

[0016] like Figure 1 As shown, the transfer application node is further divided into alarm discovery and transfer application; the rescue helicopter rushing to the transfer point node is further divided into pre-takeoff preparation, task group boarding, take-off, flight, pre-landing preparation, and landing at the transfer point; the aerial medical treatment node is further divided into transfer point take-off, flight and rescue, pre-landing preparation, and patient landing; the patient disembarkation node is further divided into patient leaving the aircraft, shutdown and inspection.

[0017] like Figure 1 As shown, the aviation emergency medical transport mission simulation module uses multi-agent simulation technology to describe the internal states of different agents and the interactive behaviors between agents, uses system dynamics technology to describe the continuous behaviors within each agent and the coordinated continuous behaviors between agents in detail, and uses Anylogic hybrid system modeling and simulation tools to visualize the simulation model.

[0018] like Figure 1 As shown, the micro-state information output by the aviation emergency medical transport mission simulation module includes rescue subject related information and patient related information; the rescue subject related information includes the time information and performance information of the pilot, random maintenance, random doctor, and random nurse staying at different participating state nodes; the patient related information includes the patient's condition change information; the macro-state information includes task related information and resource related information; the task related information includes task execution time information and task success or failure information; the resource related information includes helicopter flight route information, helicopter speed change information, and helicopter balance change information.

[0019] like Figure 1 As shown, the real-time task evaluation module includes an evaluation index system establishment and index weight confirmation module, an expert intelligent scoring module, a task intelligent evaluation module and a risk warning module; the evaluation index system establishment and index weight confirmation module: by combing through existing literature, combining actual rescue scenarios and evaluation needs, establish a general aviation emergency medical transport real-time task evaluation index system, and confirm the weight of each index; the expert intelligent scoring module: using data mining technology and artificial intelligence technology, establish an expert intelligent scoring model, and perform model training based on historical data; for receiving the micro-state information and the macro-state information transmitted by the real-time task evaluation module, selecting key feature data, simulating expert scoring behavior by comparing with historical data, and combining the evaluation index system output by the evaluation index system establishment and index weight confirmation module to score the rescue behavior status of the professional rescue team;

[0020] like Figure 1 As shown, the task intelligent evaluation module: uses the evaluation index system produced by the evaluation index system establishment and index weight confirmation module, and the scoring results corresponding to the indicators produced by the expert intelligent scoring module, selects an evaluation model, calculates the evaluation results, and uses the evaluation results for task execution effect analysis;

[0021] like Figure 1 As shown, the risk warning module is used to receive the micro-state information and the macro-state information transmitted by the real-time task evaluation module, and to receive the task evaluation result of the task intelligent evaluation module; based on historical data, for various types of professional rescue teams, establish their behavior models and learn the minimum threshold, and generate emergency warnings and response measures for each situation below the threshold.

[0022] In connection with the actual aviation emergency medical transfer scenario and taking into account the development trend of aviation medical transfer in the future, the present invention sets the scenario as a scenario in which the aviation emergency rescue force is held by a certain hospital, other hospitals make aviation medical transfer requests to the hospital, and the person in charge of the hospital decides whether to accept the request and organize the rescue force for medical transfer.

[0023] Figure 2 The figure shows a schematic diagram of the behavior prediction module of the professional rescue team. Figure 2 As shown,

[0024] Based on historical training data, the present invention proposes a rescue module time prediction model based on XGBoost with objective function optimization.

[0025] Figure 3 It is a simulation module for aviation emergency medical rescue and transport missions, such as Figure 3 As shown in the figure, the aviation emergency medical transport task simulation model is the core module of the aviation emergency medical transport real-time task evaluation system. From a microscopic perspective, the model decomposes the system into the transfer request hospital subsystem, the transfer acceptance hospital subsystem, the transfer demand subsystem and the environment subsystem. A total of 16 agents including patients, pilots, medical staff, maintenance, hospital managers, security, and environment are abstracted, and a total of 17 detailed stages of medical transport processes including patients calling for help, discovering alarms, preparations before boarding, take-off, flight, and landing are included. The discrete behavior of each agent and the collaborative behavior between different agents are realized by the ABMS method. The continuous behavior of each agent and the collaborative continuous behavior between different agents are modeled by the SD method. The aviation emergency medical transport task simulation model is characterized and realized from a microscopic perspective. Finally, the aviation medical transport task system visualization prototype is realized based on the Anylogic hybrid system modeling and simulation tool. The research results show the execution mechanism of aviation emergency medical transport tasks, reduce the cost of exercises, and lay the foundation for the research on real-time task evaluation of aviation emergency medical transport.

[0026] The input data of the model include rescue mission data (coordinates of the hospital requesting transfer and patient condition information), environmental data (weather, etc.), and professional rescue team time prediction data (possible training time information of each discrete or continuous behavior module of the professional rescue team). The output information of the model includes patient condition related information (patient treatment during transfer and patient condition information after transfer) and mission related information (transfer mission time information).

[0027] The global commitment of the aviation emergency medical transport mission system is shown in Formula 3-1, where AviMedTraModel represents the system organizational model; struct represents the organizational structure; and metaphor represents the organizational principle.

[0028] AviMedTraModel=<struct,metaphor> : (3-1)

[0029] The organizational structure of the aviation emergency medical transport mission system is shown in Formula 4-2, where Φ represents the set of intelligent subjects; Role represents the set of organizational functions; Responsibility represents the set of responsibility allocations; Goal represents the set of organizational goals; Go_relation represents the set of relationships between members who assume a certain organizational function; and Ω represents the external environment of the system.

[0030] struct=(Φ, Role, Responsibility, Goal, Go_relatio, Ω); (3-2)

[0031] Intelligent agent collection:

[0032] Φ=(A 1 , B 1 , B 2 , B 3 , B 4 , C 1 C 2 , C 3 , C 4 , C 5 , C 6 , C 7 , heli, E 1 , E 2 , E 3 ) (3-3)

[0033] Among them, A 1 Indicates patient Agent; B 1 To B 4 C represents the person in charge of the hospital, doctor, nurse and security agent who requested the transfer. 1 To C 7 The following are the person in charge of the hospital that receives the transfer, the pilot, the flight attendant, the flight attendant doctor, the flight attendant nurse, the ground flight attendant and the security agent; heli represents the helicopter agent of the hospital that receives the transfer; E 1 To E 2 They represent accident generation agent, human accident agent and environment control agent respectively.

[0034] Organizational Function Set:

[0035] Role= <R 1 .....R m >m=47 (3-4)

[0036] Among them: Help R1 , Condition Assessment 2 、Reporting of illness 3 , stretcher 4 , check patient R 5 、Patients on board 6 , Clear the field 7 , foreign matter cleaning 8 , Ground Safety Check 9 , Ground Safety Control 10 , Order restored 11 , request transfer R 12 , inter-hospital coordination 13 , In-house Command 14 , Landing Command R 15 , Task Evaluation R 16 , Plan Approval 17 , formulate a plan 18 , Mission Command 19 , Helicopter Control R 20 、Flight Accident Handling 21 , coordination and communication 22 、In-flight safety check 22 、In-flight safety observation 24 、In-machine safety control 25 , Rescue Command 26 , Coordination Signature R 27 , In-flight monitoring and treatment 28 , Rescue and Accident Handling 29 , Helicopter Inspection R 30 , Takeoff Command R 31 , observe the landmark R 32 , Safety accident handling 33 , Safety Check 34 , carrying R 35 , Transport 36 , accidentally capture R 37 , accidentally generated R 38 、Environmental Control 39 , start R 40 , Takeoff R 41 , Flight R 42 , Hover R 43 , Landing R 44 , speed change R 45 , set flight target R 46 、Patient standby preparation 47 .

[0037] Responsibility Assignment Set:

[0038] The responsibility allocation set represents the correspondence between the intelligent subject set and the organizational function set, namely:

[0039] Responsibility=f(Φ,Role) (3-5)

[0040] The set of all organizational functions that an intelligent subject participates in is the capability of the intelligent subject. The responsibility allocation set is expressed as follows:

[0041] Capacity(A 1 )= <R 1 > (3-6)

[0042] Capacity(B 1 )= <R 12 , R 13 , R 14 , R 15 > (3-7)

[0043] Capacity(B 2 )= <R 2 , R 3 , R 4 , R 5 , R 6 , R 47 > (3-8)

[0044] Capacity(B 3 )= <R 4 , R 5 , R 6 , R 47 > (3-9)

[0045] Capacity(B 4 )= <R 7 , R 8 , R 9 , R 10 , R 11 > (3-10)

[0046] Capacity(C 1 )= <R 13 , R 14 , R 15 , R 16 , R 17 > (3-11)

[0047] Capacity(C 2 )= <R 18 , R 19 , R 20 , R 21 , R 22 > (3-12)

[0048] Capacity(C3 )=<R 22 ,R 23 ,R 24 ,R 25 ,R 30 ,R 31 ,R 32 ,R 33 > (3-13)

[0049] Capacity(C 4 )=<R 4 R 5 ,R 6 ,R 22 ,R 23 ,R 24 ,R 26 ,R 27 ,R 28 ,R 29 > (3-14)

[0050] Capacity(C 5 )=<R 4 ,R 5 ,R 6 ,R 22 ,R 23 ,R 24 ,R 28 ,R 29 > (3-15)

[0051] Capacity(C 6 )=<R 11 ,R 30 ,R 31 ,R 34 > (3-16)

[0052] Capacity(C 7 )=<R 7 ,R 8 ,R 9 ,R 10 ,R 11 > (3-17)

[0053] Capacity(heli)=<R 35 ,R 36 ,R 40 ,R 41 ,R 42 ,R 43 ,R 44 ,R 45 ,R 46 > (3-18)

[0054] Capacity(E1 )= <R 38 > (3-19)

[0055] Capacity(E 2 )= <R 37 > (3-20)

[0056] Capacity(E 3 )= <R 39 > (3-21)

[0057] Organizational Goal Set:

[0058] Goal = Coordinate the efforts of multiple parties to complete the transfer task (3-22)

[0059] Mutual relationship set:

[0060]

[0061] External Environment:

[0062] Ω=(uncontrollable weather; uncontrollable accidental disasters, etc.) (3-24)

[0063] The organizational structure of the aviation emergency medical transport task system is shown in Formula 4-25, where: Know_distri_strat represents the knowledge distribution strategy; Task_distri_strat represents the task distribution strategy; Coord_control_strat represents the collaborative control strategy;.

[0064] metaphor= <Know_distri_strat,Task_distri_strat,Coord_control_strat)(3-25)

[0065] Knowledge distribution strategy:

[0066] For professional rescue teams, in order to fully reflect the impact of the technical and non-technical capabilities of the rescue subject on the execution effect of the transfer task, the model provides an interface as shown in Table 1 for inputting the personalized parameters of the rescue subject learned from the historical training data of each rescue subject. In addition, as the number of training increases, the time it takes for rescue team members to complete specific rescue tasks in the rescue process will also decrease accordingly. In order for the model to be applied to actual aviation medical transfer tasks in the future, in addition to the personalized parameters shown in the table below, a time interface is also required to input the time required for each professional rescuer to complete a specific process block in the rescue process learned from historical training data.

[0067] Table 1 Personalized parameter interface for professional aviation medical transport teams

[0068]

[0069] The real-time task of aviation emergency medical transport is represented by the geographical location of the hospital for transfer and the patient's condition. The model marks the geographical location of the transfer receiving hospital as (0, 0, 0), and the geographical location of the transfer request hospital is represented by the relative coordinates based on the transfer receiving hospital as (x, y, z); the patient's condition is represented by the remaining time t (there is no rescue value below a certain time). The real-time task can be represented as shown in Formula 3-26:

[0070] task={(x, y, z), t} (3-26)

[0071] When a transfer mission comes, it is necessary to allocate appropriate professional rescue personnel and rescue resources to the mission from the professional rescue team member pool and transfer resource pool of the transfer-accepting hospital.

[0072] The execution of aeromedical transport missions involves a variety of collaborative methods, corresponding to a variety of different collaborative control strategies:

[0073] The subject must obtain specific content instructions from a specific subject before entering the next state or performing the next task. Pure command-based collaboration is the most frequently used collaboration method in the execution of aviation medical transport tasks and is the most basic collaboration method to achieve the transport process. Pure command-based collaboration is achieved through a message passing mechanism.

[0074] The completion of a task requires the joint participation of multiple subjects, and each subject completes its own part of the task. The control strategy of task parallel collaboration is: the subject that completes the task first will enter the waiting state, and the time it takes to complete the task is the time taken by the subject that takes the longest time. When the task is completed, all participating subjects can enter other states. For example, patient safety inspection.

[0075]

[0076] The completion of a task requires the participation of multiple subjects, and each subject must keep the task synchronized at all times in order to make effective progress. For example: carrying a stretcher. The control strategy of synchronous collaboration is that the speed of all participating subjects in completing the task will be affected by the slowest subject, so the completion speed of the task is the speed of the slowest subject.

[0077]

[0078] The subject must obtain specific content instructions from a specific subject to enter the next state or perform the next task. At the same time, the status of the subject's task execution depends on certain characteristics of its collaborative subject. For example, the process of a pilot controlling a helicopter can be abstracted as giving a certain instruction to the helicopter, and the helicopter's flight status is also affected by the pilot's various ability levels. The control strategy of interactive instruction collaboration is to use the message passing mechanism to convey specific instructions, and to influence the performance of other subjects' tasks by calling the interface to pass parameters.

[0079] The task is jointly participated by multiple subjects, and each subject completes the task independently without affecting each other. The time for each subject to complete the task is different. For example, during the flight of a helicopter, all rescue subjects will conduct safety observations on the helicopter, and the subjects will not discover the same emergency at the same time. The parallel collaborative control strategy is that the time taken by the subject that takes the least time to complete the task is the final completion time of the task.

[0080]

[0081] In combination with the simulation objectives, the system needs to perform system dynamics modeling for the following continuous behaviors: position changes caused by behaviors or states (each participating entity), mentality changes (each participating entity), condition changes (patients), position changes (helicopter), propeller speed changes (helicopter), equilibrium state (helicopter), speed changes (helicopter), completion of collaborative operation tasks (stretcher carrying, safety checks, patient boarding, etc.), and use patient condition changes and helicopter control as examples to illustrate.

[0082] During air medical transfer, the patient's condition is mainly affected by three factors. On the one hand, the patient's condition will naturally deteriorate over time. On the other hand, correct and timely rescue measures by random medical staff will increase the patient's life value and delay the condition. On the other hand, various improper operations during the medical transfer process will cause further harm to the patient's condition.

[0083] Assume that the patient's condition will change with time at a speed t v0 Deterioration, when improper operation occurs, different damage values ​​will be delivered according to the degree of its impact v When a patient is being rescued, different recovery values ​​will be given to the patient depending on the level of medical care. v , t represents the cumulative life value of the patient due to time loss, hurt represents the cumulative life value of the patient due to improper operation of the professional rescue team, rescue represents the cumulative life value of the patient due to effective rescue, time represents the current remaining life value of the patient, and the patient's condition is initialized to t 0 ,but:

[0084] t=∫-t v0dt (3-30)

[0085] hurt=Σhurt v (3-31)

[0086] rescue=Σrescue v (3-32)

[0087] time=t-hurt+rescue (3-33)

[0088] The helicopter agent is a standard reactive agent. The pilot controls the helicopter state by sending commands to the helicopter, setting flight targets, changing flight parameters, etc. through the interface provided by the helicopter. The control results are mainly reflected in the changes in the helicopter displacement, propeller speed, and balance state. The changes are time-related and continuous. Therefore, when modeling the helicopter, in addition to basic agent modeling, it is also necessary to systematically characterize the helicopter flight mechanism, propeller behavior, balance state, etc.

[0089] (1) The basic parameters of the helicopter are as follows:

[0090] The initial displacement of the helicopter is expressed as: (x 0 ,y 0 , z 0 ) x 0 ,y 0 , z 0 ≥0;

[0091] Helicopter displacement representation: (x, y, z) x, y, z ≥ 0;

[0092] The displacement of the helicopter's flight target is expressed as: (t x , t y , t z ) x , t y , t z ≥0;

[0093] Pilot's planned flight speed: Speed 0 Speed 0 ≥0;

[0094] Pilot's speed control ability: Speed_control Speed_control ∈(0, 1);

[0095] The pilot's yaw control ability: Grandient_control Grandient_control ∈(0, 1);

[0096] (2) Speed ​​control modeling:

[0097] The actual flight speed of the helicopter is affected by the random factor RandomV and often deviates from the planned flight speed. 0 The pilot's speed control ability Speed_control is the degree of influence of random factors on the actual flight speed of the helicopter. When Speed_control is 1, the flight speed will not be affected by the random factors; when Speed_control is 0, the random factors will have a greater impact on the flight speed. The mathematical relationship is expressed as follows:

[0098]

[0099] The helicopter yaw behavior is the behavior of the helicopter deviating from the original route and the original flight target position. The closer the pilot's yaw control ability Grandient_control is to 1, the stronger the control ability of the flight route is, the less the flight is affected by the random factor RandomG, and the helicopter can better ensure that the flight follows the original set target; the closer Grandient_control is to 0, the weaker the control ability of the flight route is, the greater the flight is affected by the random factor RandomG, the farther the helicopter deviates from the original flight target position, and the target position after yaw is (gt x ,gt y ,gt z ) indicates that the mathematical expression of the helicopter yaw behavior is as follows:

[0100] gt x =t x +(1-Grandient_control)*RandomG (3-35)

[0101] gt y =t y +(1-Grandient_control)*RandomG (3-36)

[0102] gt z =t z +(1-Grandient_control)*RandomG (3-37)

[0103] RandomG∈R

[0104] The relative distance between the current position of the helicopter and the target position is expressed as: (x dir ,y dir , z dir ),in:

[0105] x dir =gt x-x (3-38)

[0106] y dir =gt y -y (3-39)

[0107] z dir =gt z -z (3-40)

[0108] The absolute distance between the current position of the helicopter and the target position is expressed as: dist

[0109]

[0110] The helicopter's flight velocity vector is expressed as: (v x , v y , v z ),in:

[0111]

[0112]

[0113]

[0114] The mathematical expression of helicopter displacement change is as follows:

[0115] x=x 0 +t*v x (3-45)

[0116] y=y 0 +t*v y (3-46)

[0117] x=z 0 +t*v z (3-47)

[0118] Helicopter balance banlance and pilot's balance control ability banlance control The ideal balance state Banlance should remain at 0 during the flight of the helicopter. During the flight, the balance state of the helicopter will be affected by the random factor RandomB(t) to varying degrees due to the pilot's balance control ability. RandomB(t) is a random function related to time. The mathematical expression of the balance of a helicopter is as follows:

[0119]

[0120] The propeller of a helicopter has four main states during flight: acceleration, uniform rotation, deceleration and stationary. pIndicates the propeller speed, using a p represents the acceleration of the propeller, the mathematical expression is as follows:

[0121] Acceleration state: v p =0+a p *t (3-49)

[0122] Uniform rotation state: V p =V p +0*t (3-50)

[0123] Deceleration state: v p =v p -a p *t (3-51)

[0124] Static state: v p =0 (3-52)

[0125] SD model construction includes:

[0126] Comprehensive system dynamics model of helicopter flight, speed change, target setting, yaw behavior, etc.

[0127] Figure 4 The figure shows a schematic diagram of the real-time mission assessment module for air medical transport. Figure 4 As shown in the figure, when receiving the medical transfer mission, the hospital responsible for the transfer has fully understood the patient's condition and sent professional medical staff that are suitable for his condition. The medical staff are knowledgeable and talented, and can correctly handle all possible critical situations without any operational errors. The pilots are capable and can operate the helicopter normally without any crashes. The maintenance staff are capable and will not cause safety accidents due to negligence. Although the abilities of other participants vary, their impact on the rescue operation will only be limited to time delays, and will not cause difficult-to-solve safety accidents or man-made accidents.

[0128] like Figure 4As shown in the figure, the evaluation module mainly includes two modules in order: the professional rescue team training time prediction module based on the optimized XGBoost and the aviation emergency medical transport task simulation module based on the SD-ABMS method. The professional rescue team training time prediction model in Chapter 2 can be used to predict the time information of rescuers in this transport task as the input of the aviation emergency medical transport task simulation model. This simulation model combines the time information of each rescuer in each module of the rescue process, real-time rescue task data, and task execution environment data through the simulation program, comprehensively considers the professional skills, teamwork skills, non-professional skills and other influencing factors of the rescuers, simulates the task execution in the real aviation medical transport scenario, and outputs the simulation results. The simulation results include the time information of the end of the task execution, the treatment of the patient during the transport process, and the patient's condition information at the end of the transport. Finally, the success or failure of the medical transport task is predicted based on the above simulation results. When the transport task application comes, it provides timely decision support for the relevant person in charge on whether to accept the transport task.

[0129] like Figure 4 As shown, the aviation emergency medical transport task simulation model of the general aviation emergency medical transport real-time task evaluation module is based on the aviation emergency medical transport task simulation model, assisted by the professional rescue team training time prediction model, and the training time prediction model provides the most important part of the data input for the simulation model.

[0130] The input data of the simulation model mainly include eight types of data: predictable professional rescuer-related data, professional rescuer-related statistical data, unpredictable real-time task data, unpredictable auxiliary rescuer data, planning data, environmental data and random factors.

[0131] The predictable data related to professional rescuers include: the behavior of professional rescue teams during the mission is very important to the mission execution results. The professional rescuers mentioned here include: pilots, random maintenance, random doctors, random nurses. The relevant data that can be predicted through historical training data and mission execution data are mainly the possible time usage of each rescuer in each module.

[0132] This part of data can fully reflect the proficiency of rescuers in related projects. The proficiency of rescuers in a certain project will change according to a certain trend over a period of time with their training level and task execution. There are regularities to be found, and it is not static. Since the behavior of rescuers in this rescue mission is unknown, it is necessary to predict through historical data to more realistically represent the current level of rescuers.

[0133] In order to facilitate the subsequent representation of relevant data, the professional rescue teams are first formally represented:

[0134] T={P,M,D,N} (4-1)

[0135] Among them, D represents the professional general aviation emergency medical transfer team, P represents the pilot, M represents the random maintenance personnel, D represents the random doctor, and N represents the random nurse.

[0136] Formal representation of pilot predictability-related data

[0137] Time(P) = {Tp 0 , Tp 1 ,…Tp n} n=7 (4-2)

[0138] Among them, Time (P) represents the set of relevant data that the pilot can predict, and the identifiers and meanings of each data are shown in Table 2:

[0139] Table 2 Pilot predictable related data sets

[0140]

[0141] Formal representation of random machine predictable correlation data

[0142] Time(M)={Tm 0 , Tm 1 ,…Tm n} n=15 (5-3)

[0143] Among them, Time(M) represents a random machine predictable related data set, and the data identifiers and meanings are shown in Table 3:

[0144] Table 3 Random service predictable related data sets

[0145]

[0146]

[0147] Random doctors can predict the formal representation of relevant data

[0148] Time(D)={Td 0 , Td 1 ,…Td n} n=21 (4-4)

[0149] Among them, Time(D) represents the random doctor predictable related data set, and the data identifiers and meanings are shown in Table 4:

[0150] Table 4 Random doctor predictable related data sets

[0151]

[0152]

[0153] Formal representation of random nurse predictable correlation data

[0154] Time(N) = {Tn 0 , Tn 1 ,…Tn n} n=18 (4-5)

[0155] Among them, Time(N) represents the random nurse predictable related data set, and the data identifiers and meanings are shown in Table 5:

[0156] Table 5 Random nurse predictable related data sets

[0157]

[0158]

[0159] Statistics on professional rescue workers

[0160] The statistical data related to professional rescuers mainly refers to their relevant ability data. This part of the data is generally obtained through expert assessment rather than through historical data prediction. It is presented as configurable data in the system. The ability data of professional rescuers will indirectly affect the rescue performance of the rescuers, thereby affecting the rescue effect.

[0161] Pilot statistics

[0162] Cap(P) = {Cp 0 , Cp 1 , Cp 2 …Cp n} n=6

[0163] Cp i ∈[0,1] i=0,1,…,6 (4-6)

[0164] Among them, Cap(P) represents the pilot capability set; Cp 0 Indicates speed control capability; Cp 1 Indicates balance control ability; Cp 2 Indicates yaw control capability; Cp 3 Indicates command ability; Cp 4 Indicates teamwork ability; Cp 5 Indicates emergency response capability; Cp 6 Indicates psychological endurance; the value range of various abilities is 0 to 1, and the closer to 1, the stronger the ability.

[0165] Random maintenance related statistics

[0166] Cap(M)={Cm 0 , Cm 1 , Cm 2 …Cm n} n=6

[0167] Cm i ∈[0,1] i=0,1,…,6 (4-7)

[0168] Among them, Cap(M) represents the random service capability set; Cm 0 Indicates security control capability; Cm 1 Indicates emergency response capability; Cm 2 Indicates psychological endurance; Cm 3 Indicates observation ability; Cm 4 Indicates command ability; Cm 5 Indicates helicopter inspection capability; Cm 6 Indicates the ability to collaborate. The value range of various abilities is 0 to 1. The closer to 1, the stronger the ability.

[0169] Random Doctor Related Statistics

[0170] Cap(D) = {Cd 0 , Cd 1 , Cd 2 …Cd n} n=6

[0171] Cd i ∈[0,1] i=0,1,…,6 (4-8)

[0172] Among them, Cap(D) represents the random doctor capability set; Cd 0 Indicates safety observation capability; Cd 1 Indicates command ability; Cd 2 Indicates psychological endurance; Cd 3 Indicates emergency response capability; Cd 4 Indicates the proficiency of using medical equipment; Cd 5 Indicates the level of medical skills; Cd 6 Indicates the ability to collaborate. The value range of various abilities is 0 to 1. The closer to 1, the stronger the ability.

[0173] Random nurse related statistics

[0174] Cap(N) = {Cn 0 , Cn 1 , Cn 2 …Cn n} n=5

[0175] Cn i ∈[0,1] i=0,1,…,5 (4-9)

[0176] Among them, Cap(N) represents the random nurse capability set; Cn 0 Indicates safety observation capability; Cn 1 Indicates psychological endurance; Cn 2 Indicates emergency response capability; Cn 3 Indicates the proficiency of using medical equipment; Cn 4 Indicates the level of care; Cn 5 Indicates collaborative ability; the value range of various abilities is 0 to 1, and the closer to 1, the stronger the ability.

[0177] Unpredictable real-time task data

[0178] Real-time task data is provided by the person in charge of the transfer application hospital to the person in charge of the transfer acceptance hospital when applying for the transfer task. It is related to the actual task and unpredictable. Real-time task data mainly includes the patient's condition and the geographical location of the transfer application hospital, as shown in formula 5-10:

[0179] Mission={(x, y, z), t} (4-10)

[0180] In the simulation model, the geographical location of the transfer receiving hospital is marked as (0, 0, 0), and the geographical location of the transfer requesting hospital is represented by the relative coordinates based on the transfer receiving hospital as (x, y, z); the patient's condition is represented by the remaining time t (there will be no rescue value below a specific time).

[0181] Unpredictable data for assisting rescuers

[0182] Auxiliary rescue personnel are personnel who assist professional rescue teams in medical transport during aviation medical transport. As shown in Formula 5-11:

[0183] F={GM, HP, HS, RP, RD, RN, RS} (4-11)

[0184] Among them: F means the assembly of auxiliary rescue personnel; GM means ground maintenance; HP means the person in charge of the hospital receiving the transfer; HS means the security of the hospital receiving the transfer; RP means the person in charge of the hospital requesting the transfer; RD means the doctor of the hospital requesting the transfer; RN means the nurse of the hospital requesting the transfer; RS means the security of the hospital requesting the transfer.

[0185] Ground maintenance

[0186] Time(GM) = {Tgm 0 , Tgm 1 ,…Tgm n} n=8 (4-12)

[0187] Among them, Time(GM) represents the unpredictable data set of ground maintenance, and the data identifier and its meaning are shown in Table 7:

[0188] Table 7 Ground maintenance unpredictable related data sets

[0189]

[0190]

[0191] Person in charge of the hospital receiving the transfer

[0192] Time(HP)={Thp 0} (4-13)

[0193] Among them, Time(HP) represents the unpredictable data set of the person in charge of the transfer hospital. The data identifiers and their meanings are shown in Table 8:

[0194] Table 8 Datasets related to the unpredictability of the person in charge of the transfer hospital

[0195]

[0196] Transfer to undertake hospital security

[0197] Time(HS)={Ths 0 ,Ths 1 ,…Ths n} n=4 (4-14)

[0198] Among them, Time(HS) represents the unpredictable data set of the security guards of the transfer-undertaking hospital. The data and their meanings are shown in Table 9:

[0199] Table 9 Datasets related to ground maintenance unpredictability

[0200]

[0201] Transfer request hospital doctor

[0202] Time(RD)={Trd 0 ,Trd 1 , …, Trd n} n=5 (4-15)

[0203] Among them, Time(RD) represents the unpredictable data set of hospital doctors requesting transfer. The data and their meanings are shown in Table 10:

[0204] Table 10 Datasets related to unpredictable transfer request hospital doctors

[0205]

[0206] Transfer request hospital nurse

[0207] Time(RN) = {Trn 0 , Trn 1 ,…,Trn n} n=5 (4-16)

[0208] Among them, Time(RN) represents the unpredictable data set of hospital nurses who requested transfers. The data and their meanings are shown in Table 11:

[0209] Table 11 Datasets related to unpredictable transfer request hospital nurses

[0210]

[0211] Transfer request hospital security

[0212] Time(RS) = {Trs 0 , Trs 1 ,…Trs n} n=4 (4-17)

[0213] Among them, Time(RS) represents the unpredictable data set of hospital security for transfer request, and the data and meaning are shown in Table 12:

[0214] Table 12 Datasets related to unpredictable hospital security for transfer requests

[0215]

[0216] Transfer request hospital director

[0217] The person in charge of the hospital requesting the transfer mainly plays the role of communication, coordination and internal command during the medical transfer process. If there is nothing unexpected about his coordination ability, there is no data that will have a significant impact on the transfer process.

[0218] Planning data

[0219] In this simulation system, planning-related data refers to the data that can be controlled in the mission and planned in advance. It includes flight route, flight speed and flight altitude. PR is used to represent the flight route; PV is used to represent the flight speed; PH is used to represent the flight altitude. PR can be represented as a collection of multiple points in three-dimensional space. It should be noted that all points are given according to their relative positions to the transfer receiving hospital.

[0220] PR = {(0, 0, 0), (x 1 ,y 1 , z 1 ),…,(xn ,y n , z n )} (4-18)

[0221] Environmental data

[0222] The environmental data of the general aviation emergency medical transport mission system refers to extreme weather data that may affect the mission, such as strong winds, sandstorms, heavy rains, lightning, etc. At present, the impact mechanism of the environment on the system is not considered, and the default mission environment is within the airworthiness range.

[0223] Random Factor

[0224] The random factors set in this simulation system are mainly the random factors related to the helicopter control by the pilot. They include the flight speed control random factor RaV, the yaw control random factor RaG and the balance control random factor RaB. The stronger the pilot's related control ability, the less impact the random factors have on him. If the relevant person in charge has a conservative attitude towards the task, the random factor can be set larger, otherwise it can be set smaller.

[0225] 4.1.3 Formal representation of output data

[0226] The output data of the simulation system includes the time points of key nodes in the execution of the transfer task and the changes in the patient's condition during the transfer process, which serves as the basic data for subsequent real-time task evaluation.

[0227] Key time nodes during task execution

[0228] The key time points in the task execution process are the time points of the key nodes of the task execution obtained by simulating the task execution process through the simulation model when all predictable data are input and other data are set in place. They are used to assist in the assessment of the patient's condition and provide a reference for the information exchange between departments during the execution of the transfer task. The specific time point identification and significance are shown in Table 13:

[0229] Table 13 Key time points for executing general aviation emergency medical transport missions

[0230]

[0231] Patient's condition

[0232] The patient is the party in need of transfer, represented by P. In order to conveniently describe the changes in the patient's condition during the transfer process, several important time points in the transfer process are selected to describe the patient's condition. As shown in Table 14:

[0233] Table 14 Key time points for characterizing the patient's condition

[0234]

[0235]

[0236] 4.1.4 Real-time Task Evaluation Module

[0237] The purpose of general aviation emergency medical transfer missions is to transfer critically ill patients to appropriate target hospitals in a timely manner. Therefore, for the relevant person in charge of the hospital that undertakes the transfer, the goal of mission evaluation is to see whether the patient's condition at each key node meets the conditions for successful transfer to the target hospital for a higher level of medical treatment, whether there are legal or public opinion risks due to the patient's death during the transfer process, and whether there is a waste of transfer resources because the patient is not worth transferring.

[0238] Assume that the lower limit of the patient's condition that the transfer receiving hospital can accept at each key node is:

[0239] The lower limit of the patient's condition that can be accepted by the transfer application node is t 0 ;

[0240] The lower limit of the patient's condition that can be accepted at the start node of the transfer task is t 1 ;

[0241] The lower limit of the patient's condition that can be accepted by the patient's access node is t 2 ;

[0242] The lower limit of the patient's condition that can be treated when the patient arrives at the target hospital is t 3 ;

[0243] Then, the real-time task evaluation module is:

[0244] Step 1: If Tp 0 ≤t 0 , the patient's condition no longer meets the transfer criteria and will not be accepted for transfer. 0 >t 0 , the patient's current condition meets the transfer application criteria and can continue to be evaluated. Go to step 2;

[0245] Step 2: If Tp 1 ≤t 1 , then when the transfer task begins, the patient no longer meets the transfer criteria, and continuing the transfer will cause a large waste of resources and incur a greater risk, so the transfer task is not accepted. 1 >t 1 , indicating that the patient's condition still meets the transfer criteria when the transfer mission begins, and the assessment continues. Go to step 3;

[0246] Step 3: If Tp 2 ≤t 2, the patient's condition no longer meets the criteria for boarding the aircraft, and continuing to board the aircraft will cause greater risks. In order to prevent the waste of rescue resources, the transfer mission will not be accepted. 3 >t 2 , the patient's condition still meets the criteria for computer use. Continue the assessment and proceed to step 4;

[0247] Step 4: If Tp 6 ≤t 3 , when the patient is transferred to the transfer receiving hospital, the patient's condition is already serious and no longer worth saving. Accepting the transfer task will not only incur a greater risk, but also waste rescue resources. Do not accept the transfer task. If Tp 6 >t 3 , the patient still has treatment value when arriving at the target hospital, and the transfer task is likely to be successful and has transfer value. The transfer task can be accepted.

[0248] 4.2 General Aviation Emergency Medical Transport Real-time Mission Evaluation Experiment

[0249] 4.2.1 Real-time task generation

[0250] Geographic location: A patient has an emergency and has arrived at the nearest hospital for help. The relative coordinates of the hospital relative to the geographical coordinates of Bayannur City Hospital can be converted to (30,40,0) in kilometers. That is, the hospital is 50 kilometers away from Bayannur City Hospital in Inner Mongolia.

[0251] Patient's condition: After the patient was admitted to the hospital, the doctor conducted an emergency assessment of the patient's condition and believed that the patient's condition was very serious and had exceeded the hospital's current treatment capabilities. It was recommended that the patient be transferred to Bayannur City Hospital by air emergency medical transport so that he could receive a higher level of medical care as quickly as possible. After obtaining the consent of the patient and his family, the doctor assessed the patient's current condition as being able to continue for 60 minutes, and that he would die if he did not receive effective treatment within 60 minutes. The situation was reported to the person in charge of the hospital.

[0252] Transfer application: The person in charge of the hospital urgently reports the patient's condition and the hospital's geographical location to the relevant person in charge of Bayannur City Hospital. At this point, a real-time air medical transfer task has been generated, and the task can be expressed as follows:

[0253] Mission={(30, 40, 0), 60}

[0254] 4.2.2 Input Data Preparation

[0255] After receiving the task, the person in charge of the transfer hospital conducts an emergency assessment of the task and quickly finds the currently available and suitable pilot P, random mechanic M, random doctor D and random nurse N from the professional rescue team database.

[0256] Get flight plan data

[0257] The person in charge of the transfer hospital promptly notified the pilot P and asked him to make a flight plan as soon as possible. The pilot made a flight plan and returned it to the person in charge: the flight speed PV is 300km / h; the flight altitude PH is 3000m; the flight route PR is:

[0258] (0, 0, 0) → (0, 0, 3) → (30, 40, 3) → (30, 40, 0) → (30, 40, 3) → (0, 0, 3) → (0, 0, 0)

[0259] Predictable professional rescue team data preparation

[0260] Based on their historical training data and historical task execution data, and the XGBoost rescue module time prediction model based on objective function optimization proposed in this paper, the time information of each module in this medical transport process is predicted as shown in Table 15:

[0261] Table 15 The time prediction results of each module of professional rescue team members in this mission

[0262]

[0263]

[0264] Preparation of relevant statistical data for professional rescue team members

[0265] Subsequently, the relevant person in charge retrieved the relevant capability assessment data of the four persons from the system, as shown in Table 16:

[0266] Table 16 Relevant ability assessment data of professional rescue team members

[0267]

[0268] Unpredictable data preparation for rescuers

[0269] The person in charge believes that there is nothing special about this task and that the relevant auxiliary staff of the transfer application hospital and the auxiliary staff of this hospital can complete the task at a normal level. Therefore, the data are set by default. The default data settings are shown in Table 17:

[0270] Table 17 The time data setting of auxiliary rescue personnel in this mission

[0271]

[0272]

[0273] Environmental data preparation

[0274] The person in charge checked the weather conditions along the flight route and found no extreme weather conditions, all of which were suitable for space travel.

[0275] Random Factor

[0276] Later, the person in charge was optimistic about the mission because the pilot had flown this flight route many times and there was no extreme terrain, so he set all random factors to 0.

[0277] After preparing the data, the relevant person in charge inputs the data into the simulation model.

[0278] 4.2.3 Output Data

[0279] By inputting the above data into the general aviation emergency medical transport mission simulation model, the actual simulation results obtained are as follows:

[0280] The key time nodes for task execution are shown in Table 18, in minutes:

[0281] Table 18 Key time node data during the execution of this task

[0282]

[0283] The patient's condition at key time points is shown in Table 19, in minutes:

[0284] Table 19 Patients' condition at key time points

[0285]

[0286] Real-time mission assessment

[0287] Parameter settings

[0288] The lower limit of the patient's condition that Bayannur City Hospital can accept at each key node is:

[0289] The minimum acceptable patient condition at the transfer application node is 40 minutes;

[0290] The lower limit of the acceptable patient condition at the start node of the transfer task is 35 minutes;

[0291] The lower limit of the patient's acceptable condition at the patient's on-boarding node is 20 minutes;

[0292] The minimum condition limit for patients to be in a treatable state when they arrive at the target hospital is 10 minutes;

[0293] Evaluate

[0294] According to the real-time task evaluation module, we can get:

[0295] Step 1: Tp 0 >40, the patient's current condition meets the criteria for transfer application and can continue to be evaluated;

[0296] Step 2: Tp 1 >35, indicating that the patient's condition still meets the transfer criteria when the transfer mission begins, and the assessment continues;

[0297] Step 3: Tp 3 >20, indicating that the patient meets the criteria for computer use and continues to be evaluated.

[0298] Step 4: Tp 6 >10, indicating that when the patient was finally transferred to Bayannur City Hospital, he was still worth saving and the transfer was successful.

[0299] 4.2.5 Decision support

[0300] According to the above evaluation results, this transfer mission can be successfully executed. The patient's condition was relatively dangerous when he boarded the plane. It was because of the on-board care that the patient still had good treatment value after getting off the plane. If there was no medical transfer mission, the patient would die of illness. Therefore, whether from the results of the transfer mission or from the far-reaching significance of this rescue to the patient, this rescue has treatment value. Therefore, the decision support for the relevant person in charge to accept the transfer mission is provided for their reference.

[0301] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An aviation emergency rescue real-time task evaluation system, characterized in that, it includes: A professional rescue team behavior prediction module, which is used to call the historical rescue data and historical training data of the professional rescue team, predict the behavior performance of the professional rescue team in this rescue task, generate professional rescue team behavior prediction result data, and transmit it to the medical transportation task simulation module; An aviation emergency medical transportation task simulation module, which is used to integrate real-time rescue task data, patient data, task execution environment data, and the professional rescue team behavior prediction result data, and combine the emergency situation generation and handling mechanism, cooperation mechanism, and information transmission mechanism inside the aviation emergency medical transportation task simulation module to run the medical transportation task simulation model, obtain the micro-state information of each state node of the internal personnel of the system during the medical transportation process and the macro-state information of the medical transportation action, and transmit the micro-state information and the macro-state information to the real-time task evaluation module; A real-time task evaluation module, which calls the micro-state information and the macro-state information generated by the aviation emergency medical transportation task simulation module, uses an expert intelligent scoring model based on data mining and a well-established evaluation mechanism to generate an evaluation result of the rescue task execution effect, and provides decision-making support for the responsible personnel in the system; at the same time, it integrates the micro-state information, the macro-state information and the rescue task execution effect evaluation result, uses a risk warning model to identify the possible risks during the task execution process, and provides countermeasures; wherein, the aviation emergency medical transportation task simulation module realizes the description of the internal states of different Agents and the interaction behaviors between Agents through multi-agent simulation technology, details the continuous behaviors inside each Agent and the collaborative continuous behaviors between Agents through system dynamics technology, and realizes the visualization of the simulation model through the Anylogic hybrid system modeling and simulation tool; the aviation emergency medical transportation task simulation module uses multi-agent simulation to abstract into a transfer request hospital subsystem, a transfer receiving hospital subsystem, a transfer demand subsystem and an environment subsystem; the transfer request hospital subsystem includes a transfer request hospital building entity, a transfer request hospital lawn entity, a transfer request hospital responsible person Agent, a transfer request hospital security Agent, a doctor Agent, and a nurse Agent; the transfer receiving hospital subsystem includes a transfer receiving hospital building entity, a transfer request hospital lawn entity, a stretcher object, an airborne equipment object, a take-off battery object, a transfer request hospital responsible person Agent, a transfer request hospital security Agent, a random doctor Agent, a random doctor Agent, a random nurse Agent, a helicopter Agent, a ground maintenance Agent, a random maintenance Agent and a pilot Agent; the transfer demand subsystem includes a patient Agent; the environment subsystem includes an accident generation Agent, a human accident Agent and an environment control Agent; The micro-state information output by the aviation emergency medical transportation task simulation module includes information related to the rescue subjects and information related to the patients; the information related to the rescue subjects includes the time information and performance information of the pilot, on-board mechanic, on-board doctor, and on-board nurse staying at different state nodes; the information related to the patients includes the information on the change of the patient's condition; the macro-state information includes information related to the task and information related to resources; the information related to the task includes task execution time information and task success or failure information; the information related to resources includes helicopter flight route information, helicopter speed change information, and helicopter balance change information; The real-time task evaluation module includes an evaluation index system establishment and index weight confirmation module, an expert intelligent scoring module, a task intelligent evaluation module, and a risk warning module; the evaluation index system establishment and index weight confirmation module establishes a real-time task evaluation index system for general aviation emergency medical transportation by sorting out existing literature, combining actual rescue scenarios and evaluation requirements, and confirms the weights of each index; the expert intelligent scoring module uses data mining and artificial intelligence to establish an expert intelligent scoring model, trains the model based on historical data, receives the micro-state information and the macro-state information transmitted by the real-time task evaluation module, selects key feature data, simulates the expert scoring behavior through comparison with historical data, and scores the rescue behavior status of the professional rescue team in combination with the evaluation index system produced by the evaluation index system establishment and index weight confirmation module; the task intelligent evaluation module uses the evaluation index system produced by the evaluation index system establishment and index weight confirmation module and the scoring results corresponding to the indexes produced by the expert intelligent scoring module, selects an evaluation model, calculates the evaluation result, and uses the evaluation result for the analysis of the task execution effect; The risk warning module is used to receive the micro-state information and the macro-state information transmitted by the real-time task evaluation module, and is used to receive the task evaluation result of the task intelligent evaluation module; based on historical data, for various types of professional rescue teams, establish their behavior models and learn the minimum thresholds, and generate emergency warnings and response measures for each situation below the threshold.

2. The aviation emergency rescue real-time task evaluation system according to claim 1, characterized in that The internal personnel of the system include: personnel of the transfer-receiving hospital, personnel of the transfer-requesting hospital, and patients; the personnel of the transfer-receiving hospital include the person in charge of the transfer-receiving hospital, ground support forces, and professional rescue teams: the ground support forces include security guards of the transfer-receiving hospital and ground mechanics; the professional rescue teams include pilots, on-board mechanics, on-board doctors, and on-board nurses; the personnel of the transfer-requesting hospital include the person in charge of the transfer-requesting hospital, security guards of the transfer-requesting hospital, and medical staff.

3. The aviation emergency rescue real-time task evaluation system according to claim 1, characterized in that The historical training data includes skills training data and non-skills training data, and the skills training data includes general knowledge skills data, professional skills data, and process proficiency training data; The non-skills training data includes communication and coordination ability, command ability, psychological stress, and emergency response ability training data.

4. The real-time task evaluation system for aviation emergency rescue according to claim 1, characterized in that, The aviation emergency medical transport task simulation module divides the medical transport process into the following state nodes: transport application, linkage mechanism activation, rescue helicopter heading to the transport point, patient boarding, in-air medical treatment, patient disembarking, and in-hospital area medical treatment.

5. The real-time task evaluation system for aviation emergency rescue according to claim 4, characterized in that, The transport application node is further divided into police situation discovery and transport application; the rescue helicopter heading to the transport point node is further divided into pre-takeoff preparation, task team boarding, takeoff, flight, pre-landing preparation, and landing at the transport point; the in-air medical treatment node is further divided into takeoff from the transport point, flight and medical care, pre-landing preparation, and patient landing; the patient disembarking node is further divided into patient leaving the aircraft, powering off and inspection.

Images