CN105205565A - Controller workload prediction method and system based on multiple regression model - Google Patents
Controller workload prediction method and system based on multiple regression model Download PDFInfo
- Publication number
- CN105205565A CN105205565A CN201510645215.3A CN201510645215A CN105205565A CN 105205565 A CN105205565 A CN 105205565A CN 201510645215 A CN201510645215 A CN 201510645215A CN 105205565 A CN105205565 A CN 105205565A
- Authority
- CN
- China
- Prior art keywords
- sector
- controller
- aircraft
- indexes
- workload
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000001514 detection method Methods 0.000 claims description 68
- 238000012417 linear regression Methods 0.000 claims description 34
- 230000008859 change Effects 0.000 claims description 24
- 230000006870 function Effects 0.000 claims description 24
- 238000006243 chemical reaction Methods 0.000 claims description 22
- 230000009194 climbing Effects 0.000 claims description 15
- 230000001419 dependent effect Effects 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 11
- 238000010606 normalization Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 8
- 238000009826 distribution Methods 0.000 claims description 7
- 238000000528 statistical test Methods 0.000 claims description 6
- 230000008676 import Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims 2
- 238000004891 communication Methods 0.000 description 19
- 238000010586 diagram Methods 0.000 description 11
- 238000011160 research Methods 0.000 description 11
- 238000007726 management method Methods 0.000 description 10
- 238000012545 processing Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 230000004424 eye movement Effects 0.000 description 7
- 230000003111 delayed effect Effects 0.000 description 6
- 238000005259 measurement Methods 0.000 description 5
- 238000012887 quadratic function Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012886 linear function Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000000611 regression analysis Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 230000006399 behavior Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 230000001105 regulatory effect Effects 0.000 description 3
- 238000007619 statistical method Methods 0.000 description 3
- 238000012353 t test Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000003340 mental effect Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000004445 quantitative analysis Methods 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 206010040914 Skin reaction Diseases 0.000 description 1
- SAQSTQBVENFSKT-UHFFFAOYSA-M TCA-sodium Chemical compound [Na+].[O-]C(=O)C(Cl)(Cl)Cl SAQSTQBVENFSKT-UHFFFAOYSA-M 0.000 description 1
- 238000011497 Univariate linear regression Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000036772 blood pressure Effects 0.000 description 1
- 210000001124 body fluid Anatomy 0.000 description 1
- 239000010839 body fluid Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000556 factor analysis Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 230000035483 skin reaction Effects 0.000 description 1
- 231100000430 skin reaction Toxicity 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention discloses a controller workload prediction method and system. The controller workload prediction method comprises the steps of building a multiple regression model according to controller workload prediction related indexes and controller workload prediction sample data; importing air traffic flow situation index prediction data in a controlled sector into the multiple regression model, and obtaining a controller workload prediction result. According to the controller workload prediction method and system, the air traffic flow situation multidimensional indexes which influence controller workload are considered completely and comprehensively, and thereby effective prediction for controller workload is achieved. The designed controller workload prediction system can be applied to an engineering unit and has great operability.
Description
Technical Field
The invention relates to the field of monitoring, in particular to a method and a system for predicting the workload of a manager.
Background
With the development of the air transportation industry, in order to ensure the safety and the order of various flight activities, the air traffic control service is developed and perfected at the same time, and the air traffic control service is mature in the 20 th century and the 80 th era. The main content of modern air traffic control services is: an air traffic controller (referred to as a 'controller', the same below) manages and controls the administered aircrafts by relying on modern communication, navigation and monitoring technologies, coordinates and guides the movement paths and modes of the administered aircrafts, so as to prevent the aircrafts from colliding with the aircrafts and the aircrafts from colliding with obstacles in an airport maneuvering area, maintain and accelerate the ordered flow of air traffic. The performer of this task is an air traffic controller (referred to as a "controller," hereinafter). The controller mainly works by closely monitoring the flight dynamics through real-time information displayed by a radar and issuing various instructions to a unit through radio communication equipment, and the controller is a work integrating multiple senses such as eyes, hands, mouths and the like and coordinating together. In the busiest time period, an individual controller needs to control dozens of aircrafts at the same time, so that the controller is high in mental labor intensity and workload, frequently needs to shift, fatigue states of different levels are easy to occur, the workload of the controller is predicted, and the time period in which overload is possibly generated is interfered, so that the fatigue states can be effectively avoided.
Patent document CN104636890A discloses a method for measuring workload of air traffic controllers in 2015, 05 and 20. The method comprises the following steps: determining a control load measurement index, wherein the control load measurement index comprises an eye movement index and a voice index; and B: recording eye movement index data corresponding to each eye movement index and voice index data corresponding to each voice index in real time; and C: performing factor analysis on the recorded eye movement index data, and calculating an eye movement comprehensive factor of the eye movement index data; step D: and establishing a control load regression model by taking the eye movement comprehensive factor and the voice index as input factors and the control workload value as output factors.
Patent document CN102306297A discloses a method for measuring workload of air traffic controllers in 2012, month 01 and 04. The method comprises the steps of classifying basic air traffic events, determining classification of basic control behaviors based on human ergonomics, performing statistical analysis and control communication through a radar voice recorder, establishing a controller workload statistical model easy to observe, obtaining basic control unit workload determined through the air traffic events through machine learning, determining a correction coefficient of the workload, correcting the controller workload statistical model easy to observe by using the coefficient, and determining a controller workload measurement model. The method can realize accurate quantitative measurement of the workload of the controllers, and provides a basis for civil aviation control skill training, airspace sector capacity assessment and airspace planning design.
Research on the prediction of the workload of the controllers is mainly reflected in the evaluation technology of the workload of the controllers, and three types of methods for evaluating the workload of the controllers have evolved continuously since the 70 s of the 20 th century, namely:
(1) and (4) analyzing according to the physiological and behavior characteristics of the controller to obtain the control workload intensity. The measured physiological indexes comprise electric shock skin reaction, heart rate, electrocardiogram, blood pressure, body fluid and the like, and the behavior indexes comprise equipment operation times, air-ground communication time records and the like.
(2) Subjective evaluation methods in the form of observation and questionnaires, such as the ATWIT technique (airworthworkloadinputputtechnique, the air traffic load input technique of the federal aviation administration in the united states), the NASA-TLX scale (taskloadindex, the mission load scale of the national aviation and space administration in the united states), the SWAT scale (subjectiveworkloadanalysistechnique, the subjective workload analysis technique), and the MCH method (modifiedcoopper-harassing, kubo-sabai correction), and the like.
(3) And (3) subdividing the work of the controller, measuring the consumed time of the visible work, and converting the invisible work into the consumed time, so as to realize the quantitative evaluation of the work load of the controller in a time measurement mode. Such methods include the ICAO recommended DORATASK method (Directorite of operation research and Nalysemploying university, proposed by the British research and analysis society) and the MBB method (Messerchmidt, Bglkowand Blohmof Germany, proposed by Meissensch Schmidt, Del and bloom), and the RAMS method (Re-organiszenaTCMaterial Simulator, proposed by the European empty tube laboratory centre).
The related research content of the workload prediction of the current controllers mainly has the following defects: (1) in the aspect of research methods, more qualitative researches and less quantitative researches result in insufficient objectivity. (2) In the aspect of index research, the method starts from indexes which directly reflect the workload of the controllers, considers less influence factor indexes of the workload of the controllers, and has the advantages of single index dimension, incomplete comprehensiveness, integration and low prediction reliability. (3) In the aspect of applicability, the existing research still stays in the laboratory research stage, mainly serves strategic decisions, and has less practical engineering application for air traffic control units. Due to the defects, the current research on the workload prediction of controllers at home and abroad is lack in objectivity, comprehensiveness, accuracy, operability and the like.
Disclosure of Invention
The invention provides a controller workload prediction method and a system which are more efficient and can improve objectivity and prediction accuracy.
The purpose of the invention is realized by the following technical scheme:
a method for administrator workload prediction comprising the steps of:
step 1: selecting air traffic flow situation indexes of a control sector at a certain time interval and controller workload indexes corresponding to the situation indexes as sample data;
step 2: establishing a linear regression model and a nonlinear regression model according to the sample data;
and step 3: comparing the linear regression model with the nonlinear regression model through the fitting degree, the significance and the error analysis to determine a multiple regression model for predicting the workload of the controller;
and 4, step 4: and importing the real-time control sector air traffic flow situation indexes into a controller workload prediction multiple regression model to obtain controller workload indexes.
Further, the indexes of the air traffic flow situation of the control sector in the step 1 include a sector operation trafficability index, a sector operation complexity index, a sector operation safety index and a sector operation economy index.
Further, the control sector air traffic flow situation indexes comprise sector operation trafficability indexes, sector operation complexity indexes, sector operation safety indexes and sector operation economy indexes;
the sector trafficability detection indexes respectively comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density;
the sector complexity detection indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft;
the sector safety detection indexes comprise sector short-term conflict alarm frequency and sector minimum safety height alarm frequency;
the sector economy detection indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time.
Further, in the step 2, standard conversion is performed on the sample data; the normalization conversion process is as follows:
let xij、x′ijRespectively representing the raw data of the ith sample and the normalized and converted data,sjrespectively representing the mean and variance of the jth index data, then:
further, the step 2 specifically includes:
according to the standardized sample data x'ij(i 1,2, … m, j 1,2, … n), establishing a multiple linear regression model and a multiple non-linear regression model respectively, and solving coefficientsbi，
Wherein the multiple linear regression model is:
y ═ XB + U (formula 1)
Wherein,
the multiple nonlinear regression model is:
Y＝f[(b1,b2,…,bk)；X1,X2,…,Xn](formula 2)
The method comprises the following steps that a dependent variable Y is a controller load index, an independent variable X is an air traffic flow situation index of n control sectors, m represents a control sector operation performance index sample under m groups of time intervals, U is a random error except the influence of m independent variables on the dependent variable Y and obeys normal distribution, and f represents a nonlinear regression function.
Further, the step 3 specifically includes:
from the return of the coefficients of the decision R of each model2And (3) value, F test and t test, respectively verifying and comparing the fitting degree and the significance of the two regression models, calculating the detection errors of the two regression models on the basis of higher model fitting degree and obvious significance, and selecting the model with the minimum error as the multiple regression model for the workload prediction of the controllers.
Further, the real-time control sector air traffic flow situation indexes in the step 4 need to be subjected to standardized conversion before being input into the multiple regression model; the normalization conversion process is as follows:
mean value of n indexes of sample data according to m groups of time intervalsVariance sjFor the air traffic flow situation index t of the control sectorj(j ═ 1, 2.. times, n) normalized conversion:converting the data tj' import into controller workload prediction multiple regression model.
Further, the method includes a step 5 of responding to an alarm by the controller workload when the controller workload index exceeds a threshold.
A system for forecasting administrator workload, comprising: a component module: selecting air traffic flow situation indexes of a control sector, substituting controller workload prediction sample data corresponding to relevant indexes into a multiple linear regression model and a multiple nonlinear regression model for fitting; obtaining parameter estimation values of a multiple linear regression model and a multiple nonlinear regression model; calculating a detection error through statistical test, and determining a multiple regression model for predicting the workload of the controller; a prediction module: and importing the air traffic flow situation index prediction data of the control sector into a controller workload prediction multiple regression model to obtain a prediction result of the controller workload.
Further, the system further comprises a normalization conversion module: the system is used for carrying out standardized conversion on the sample data and the real-time data of the air traffic flow situation indexes of the control sector; an alarm module: when the prediction result exceeds the threshold value, the controller workload responds to the alarm.
Furthermore, the prediction system also comprises a control sector traffic flow situation detection database, a data leading device and an index acquisition device which are coupled with the control sector traffic flow situation detection database;
the data leading device comprises a telegraph data interface, a comprehensive track data interface and a control voice data interface which are respectively coupled with the control sector traffic flow situation detection database;
the index acquisition device is used for acquiring air traffic flow situation indexes of a control sector, wherein the air traffic flow situation indexes of the control sector comprise a sector operation trafficability index, a sector operation complexity index, a sector operation safety index and a sector operation economy index; the sector trafficability detection indexes respectively comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; the sector complexity detection indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft; the sector safety detection indexes comprise sector short-term conflict alarm frequency and sector minimum safety height alarm frequency; the sector economy detection indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time;
the component module reads the related indexes of the controller workload prediction and the sample data of the controller workload prediction from the traffic flow situation detection database of the control sector; and the prediction module reads the air traffic flow situation index prediction data of the control sector from the traffic flow situation index detection database of the control sector.
The invention has the beneficial effects that:
the method adopts a quantitative analysis method, calculates accurate air traffic flow situation index data in a future time period by uninterrupted detection and calculation analysis of mass operation data, acquires the relation between the air traffic flow situation and the workload of a controller by mining historical data, predicts the workload of the controller on the basis, has the advantages of objectivity, high efficiency and accuracy, and avoids the defect of empirical management such as easy fatigue and easy subjectivity of manual prediction. More importantly, the system can meet the actual requirements of air traffic control units on real-time prediction and warning of workload of controllers, and has a data support effect on improving the control operation management level and optimizing the control airspace structure. The invention comprehensively and comprehensively considers the multi-dimensional indexes of the air traffic flow situation influencing the workload of the controller, thereby realizing the effective prediction of the workload of the controller. The designed system for predicting the workload of the controllers can be applied to engineering units and has strong operability.
Drawings
FIG. 1 is a diagram illustrating a method for predicting workload of a controller according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a controller workload prediction system according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a system logic structure for workload prediction of a second controller according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a system network structure for workload prediction of controllers according to an embodiment of the present invention;
FIG. 5 is a functional block diagram of a system for workload prediction for controllers according to an embodiment of the present invention;
FIG. 6 is a schematic structural diagram of a third integrated track data acquisition function according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of a voice data collection process according to a third embodiment of the present invention;
fig. 8 is a schematic structural diagram of a data acquisition function of a third telegram according to an embodiment of the invention;
FIG. 9 is a schematic diagram of a method for four-supervisor workload prediction in accordance with an embodiment of the present invention;
FIG. 10 is a graph showing the results of a four-fold nonlinear regression fit in accordance with an embodiment of the present invention;
FIG. 11 is a diagram illustrating error of a four-fold nonlinear regression fit according to an embodiment of the present invention;
FIG. 12 is a schematic structural diagram of a five-controller workload prediction system according to an embodiment of the present invention;
wherein: 1. a component module; 2. a prediction module; 3. a standardized conversion module; 4. an alarm module: 5. operating an index detection database; 6. a data leading device; 7. index collection system.
Detailed Description
The invention is further described with reference to the drawings and the preferred embodiments.
Example one
As shown in fig. 1, the method for predicting the workload of the controller disclosed in the present embodiment includes the steps of:
s1, selecting air traffic flow situation indexes of control sectors, substituting control staff workload indexes corresponding to relevant indexes as sample data into a multiple linear regression model and a multiple nonlinear regression model for fitting; obtaining multiple linear regression models and multiple nonlinear regression model parameter estimation values, and establishing multiple linear regression models and multiple nonlinear regression models;
s2, comparing the linear regression model with the nonlinear regression model through fitting degree, significance and error analysis, and determining a controller workload prediction multiple regression model;
and S3, importing the real-time data of the air traffic flow situation indexes of the control sector into the controller workload prediction multiple regression model to obtain the prediction result of the controller workload index.
As shown in fig. 2, the present embodiment also discloses a system for forecasting the workload of an administrator, comprising,
a component module: selecting a controller workload prediction related index, and substituting controller workload prediction sample data corresponding to the related index into the multiple regression model for fitting; obtaining a parameter estimation value and a sample output value of the multiple regression model; importing the parameter estimation value and the sample output value into a multiple regression model to obtain a sample regression function;
a prediction module: and importing the prediction data of the air traffic flow situation indexes of the control sector into a sample regression function to obtain a prediction result of the workload of the controller.
Regression analysis is an important branch of multivariate statistical analysis, which is a statistical method for detecting one or more response variables (i.e., dependent variables) through a set of detection variables (i.e., independent variables). The case of only one dependent variable is called univariate regression, and a plurality of dependent variables is called multivariate regression. Considering that the workload of the controller is influenced by various factors, the workload of the controller is set as a single response variable, so that a unitary multiple regression method (multiple regression for short) is adopted to predict the workload of the controller.
According to the linear relationship of the regression function, two basic function models, namely multiple linear regression and multiple nonlinear regression, can be classified. The invention can adopt two models to be used, then selects one with small error as the final prediction model, and can also select one for prediction to simplify the operation process.
The method adopts a quantitative analysis method, calculates accurate air traffic flow situation index data in a future time period by uninterrupted detection and calculation analysis of mass operation data, acquires the relation between the air traffic flow situation and the workload of a controller by mining historical data, predicts the workload of the controller on the basis, has the advantages of objectivity, high efficiency and accuracy, and avoids the defect of empirical management such as easy fatigue and easy subjectivity of manual prediction. More importantly, the system can meet the actual requirements of air traffic control units on real-time prediction and warning of workload of controllers, and has a data support effect on improving the control operation management level and optimizing the control airspace structure. The invention comprehensively and comprehensively considers the multi-dimensional indexes of the air traffic flow situation influencing the workload of the controller, thereby realizing the effective prediction of the workload of the controller. The designed system for predicting the workload of the controllers can be applied to engineering units and has strong operability.
Example two
The embodiment discloses a system architecture which is used as an implementation platform of the controller workload predictor system and can be used for implementing the prediction method.
Fig. 3 shows a structure of a controller workload prediction system according to the present embodiment. The air traffic controller workload prediction system mainly comprises a set of control sector traffic flow situation detection database and three functional modules of data leading and index acquisition. The traffic flow situation detection database of the control sector classifies and stores the air traffic flow situation data (including radar comprehensive flight path data, telegraph data, VHF recording data and the like) collected by each information collection point, and provides data basis for the work load prediction of a controller.
Fig. 4 and 5 disclose a network structure and a corresponding functional module structure for implementing the prediction system of the present invention. The system collects real-time data through a data acquisition server, monitors operation data in real time through a control sector traffic flow situation detection server and a controller workload prediction and alarm server, predicts the controller workload in a future time period, and alarms in the time period when the workload exceeds a threshold value. The network platform of the whole system is physically isolated from the air traffic control production network by depending on the existing management information network, the acquisition platform and the air traffic control production network, so that the one-way transmission of data is ensured, the network attack is prevented, and the safety of related data and the reliability of a production operation system are ensured.
EXAMPLE III
The embodiment discloses a management operation data acquisition scheme, which comprises but is not limited to acquisition of relevant indexes of controller workload prediction, sample data of controller workload prediction and prediction data of air traffic flow situation indexes of a control sector.
The study uses the controller workload index as a dependent variable and is marked as Y. The total number of the air traffic flow situation indexes of the control sector is 15, and the independent variable X is recorded as follows:
X＝{Xii ═ 1,2, …,15} (formula 3.1)
Wherein the sector trafficability detection index is { X }1,X2,X3,X4Respectively representing sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; sector complexity detection index is { X }5,X6,X7,X8Respectively representing the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed changing times of the sector aircraft and the navigation changing times of the sector aircraft; sector security detection index is { X9,X10Respectively representing a sector short-term conflict alarm frequency and a sector minimum safe height alarm frequency; sector economy detection index is { X11,X12,X13,X14,X15And the sector saturation, the sector queuing length, the sector aircraft delay frame rate, the sector aircraft delay time and the sector aircraft average delay time are respectively represented. The parameter indexes are mainly acquired from the following aspects.
Integrated track acquisition
The air traffic control automation system performs data fusion and data processing on monitoring signals of a first air traffic control radar, a second air traffic control radar and the like, outputs comprehensive flight path information, and mainly comprises a radar front-end processing module, a radar data processing module and a flight plan processing module.
According to the technical scheme, the air traffic control automation system collects the comprehensive track data and transmits the comprehensive track data in a network mode. And the data acquisition server analyzes the acquired comprehensive track data, and acquires information such as the height, the speed and the position of the aircraft for index calculation.
The integrated track data acquisition module comprises a track data format conversion module, a track data analysis module and a track data storage module, as shown in fig. 6.
Voice data collection
The controller and the pilot realize the land-air voice communication through the VHF communication system. The system consists of a Very High Frequency (VHF) transceiver station and a signal transmission and processing device.
The voice data acquisition is connected with the voice signal acquisition from the distribution frame in parallel, and the land-air communication information is decoded and stored for the analysis of the traffic load controlled and directed by the controller.
As shown in fig. 7, seat voice data is introduced into the system data acquisition server from the internal telephone system distribution frame through the shielded network cable, and the voice channel corresponds to the seat (sector).
The voice signal is from the high impedance (the recording module is 200K ohm) collection (controller ground-to-air conversation) seat pronunciation on the distribution frame, does not influence ground-to-air conversation and voice recording, adopts many pairs of cable wires to lead the voice signal from the distribution frame to the voice processor, realizes the collection and the analysis to a plurality of seat pronunciation.
Telegraph data acquisition
The telegraph forwarding system is a junction device for receiving and transmitting civil aviation flight dynamic fixed telegraph, and messages of the civil aviation flight dynamic fixed telegraph are formed by arranging a plurality of specified data groups according to a fixed sequence.
The telegram data acquisition module is used for leading telegram data output by the telegram forwarding system, carrying out format conversion, analysis and storage on the data, and acquiring flight plan data, as shown in fig. 8. The module analyzes the received telegram data and stores the analyzed telegram data in a database for storage, and the data is used for calculating the operation performance index of the sector.
Control sector air traffic flow situation index collection
The system collects real-time operation data such as comprehensive flight path, flight plan, voice communication and the like from an air traffic control automation system, a retransmission system and a telephone system, and establishes an air traffic flow situation index system of a control sector by taking relevant documents of the international civil aviation organization (ICAO for short, the same below) and the Federal Aviation Administration (FAA) as references, wherein the index system comprises the following components: the sector operation trafficability indexes comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; the sector operation complexity indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft; and the sector economic indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time. And outputting the detection result of the air traffic flow situation indexes of the control sector based on the index system. The system provides a good human-computer interface for a user to check various real-time statistical graphs.
Sector operation trafficability index
(1) Sector traffic
Sector traffic refers to the number of aircraft frames administered per unit time by the regulatory sector. The system acquires the position information of the air aircraft by leading the air traffic control automation system to synthesize track data, and calculates the sector flow by combining the configured sector boundary information.
(2) Sector voyage mileage
The sector navigation mileage refers to the sum of the aircraft navigation mileage governed by the control sector per unit time. Setting the number of times of the aircraft frames in unit time of the control sector as n and the navigation mileage of the jth aircraft as MiThe navigation mileage of the sector is MtotalThen, thenThe system acquires the position information of the air aircraft by leading the air traffic control automation system to synthesize track data, and calculates the sector navigation mileage by combining with the configured sector boundary information.
(3) Sector voyage time
The sector navigation time refers to the unit time of the control sectorThe sum of the administered aircraft voyages. Setting the number of times of aircraft frames in unit time of a control sector as n and the navigation time of the ith aircraft as TiThe sector navigation time is TtotalThen, thenThe system acquires the position information of the air aircraft by leading the air traffic control automation system to synthesize track data, and calculates the sector flight time by combining the configured sector boundary information.
(4) Sector traffic flow density
Sector traffic flow density is a measure of how dense the aircraft is in the jurisdiction of a control sector per unit time. Let the sector area be SsecThe sector flow in unit time is n, and the sector traffic flow density in unit time is DsecThen D issec＝n/Ssec. The system reads the configured sector boundary information to obtain the sector area, and the sector traffic flow density is obtained by combining the sector flow calculation.
Sector operation complexity indicator
(1) Sector number of aircraft climbs
The sector aircraft climbing times refer to the sum of the aircraft climbing times governed in unit time of the control sector. Setting the number of times of aircraft frames in unit time of a control sector as n and the number of times of climbing of the ith aircraft as ciThe number of times of climb of the sector aircraft is ctotalThen, thenThe system is connected with the track data in real time in a comprehensive mode, the climbing condition of aircrafts in the sector is monitored and counted, one aircraft climbs one altitude layer for one time, and the climbing times of the aircrafts in the sector are calculated.
(2) Sector aircraft descent times
Sector aircraft descent times refers to a regulatory sectorThe sum of the number of aircraft drops per unit time. Setting the number of times of aircraft frames in unit time of a control sector as n and the number of times of descent of the ith aircraft as DiSector number of aircraft descent DtotalThen, thenThe system is connected with the track data in real time in a leading mode, descending conditions of aircrafts in the sectors are monitored and counted, one aircraft descends for one time in one altitude layer, and the number of climbing times of the aircrafts in the sectors is calculated.
(3) Sector aircraft speed change times
The number of times the aircraft changes speed in a sector refers to the sum of the number of times the aircraft speed changes per unit time in the regulatory sector. Setting the number of times of aircraft frames in unit time of a control sector as n and the number of times of speed change of the ith aircraft as SiThe number of speed changes of the sector aircraft is StotalThen, thenThe system is connected with the real-time comprehensive track data, the speed change condition of the aircrafts in the sector is monitored and counted, the speed of one aircraft continuously changes to reach the set parameter of one speed change, and the speed change times of the aircrafts in the sector are obtained through calculation.
(4) Sector number of aircraft diversions
The sector aircraft diversion times are the sum of the aircraft heading change times in the unit time of the control sector. Setting the number of times of aircraft frames in unit time of a control sector as n and the number of times of flight change of the ith aircraft as HiSector aircraft number of changes is HtotalThen, thenThe system is connected with the real-time comprehensive track data, the course change condition of the aircrafts in the sector is monitored and counted, the course of one aircraft continuously changes until the set parameter is changed into a course change, and the course change is calculated to obtainSector number of aircraft diversions.
1.1.1.1.1 sector operational safety index
(1) Sector short term collision warning frequency
The sector short-term conflict warning frequency refers to the number of aircraft short-term conflict warnings administered in unit time of a control sector, and is obtained through STCA warning data statistics of a system leading and connecting an air traffic control automation system.
(2) Minimum safe altitude warning frequency of sector
The alarm frequency of the lowest safe altitude of the sector refers to the alarm times of the lowest safe altitude of the aircraft administered in unit time of the control sector, and is obtained by MSAW alarm data statistics of a system leading and receiving air traffic control automation system.
1.1.1.1.2 sector operating economy index
(1) Sector saturation
The sector saturation is the ratio of the flow rate to the capacity of the control sector in unit time, and the maximum number of aircrafts which can be administered in unit time of the control sector is calibrated to be the capacity of the control sector. Setting the aircraft frame number of times in unit time of the control sector as n, the capacity of the control sector as C, and the saturation of the sector as SatusecThen Satusecn/C. And the system reads the configured sector capacity parameters and obtains the sector saturation by combining the sector flow calculation.
(2) Sector queue length
In the aircraft administered in the unit time of the control sector, if a queuing condition such as spiral waiting occurs when entering the sector, the aircraft is defined as a queuing aircraft, and the sector queuing length is defined as the number of the queuing aircraft. And leading the system to connect the comprehensive track data, judging whether the target aircraft is in circling waiting at the boundary of the sector, and calculating to obtain the sector queuing length.
(3) Sector aircraft delay rate
Under control of pipeAmong the aircrafts governed in the sector unit time, the aircraft with the voyage time exceeding the normal range is defined as a delay aircraft, and the part of the voyage time exceeding the normal range is defined as the delay time. Setting the number of aircraft frames in unit time of a control sector as n, the number of delayed frames of the sector aircraft as d, and the rate of delayed frames of the sector aircraft as landsecThen at is just drivesecD/n. And leading the comprehensive track data by the system, comparing the actual flight time of each aircraft in the control sector with the empirical flight time, if the actual flight time is greater than the empirical flight time, determining that the aircraft is delayed, and calculating to obtain the delayed frame rate of the aircraft in the sector.
(4) Sector aircraft latency
Among aircrafts governed by a control sector within unit time, an aircraft with the voyage time exceeding a normal range is defined as a delay aircraft, the part of the voyage time exceeding the normal range is defined as delay time, and the sum of the delay time is defined as the delay time of the aircraft in the sector. Setting the number of times of aircraft frames in unit time of a control sector as n and the Delay time of the ith aircraft as DelayiDelay time of aircraft in sectorsecThen, thenAnd leading the comprehensive track data by the system, comparing the actual flight time of each aircraft in the control sector with the empirical flight time, if the actual flight time is greater than the empirical flight time, determining that the aircraft is delayed, and calculating to obtain the delay time of the aircraft in the sector.
(5) Sector aircraft mean time delay
Among the aircrafts governed by the control sector in unit time, the aircraft with the voyage time exceeding the normal range is defined as a delay aircraft, and the part of the voyage time exceeding the normal range is defined as the delay time. Setting Delay time of sector aircraft as DelaysecThe number of aircraft frames in the unit time of the control sector is n, and the average delay time of the aircraft with the sector isDavgsecThen Davgsec＝DelaysecAnd/n. And leading the comprehensive flight path data by the system, comparing the actual flight time of each aircraft in the control sector with the empirical flight time, if the actual flight time is greater than the empirical flight time, determining that the aircraft is delayed, and calculating to obtain the average delay time of the aircraft in the sector.
Controller workload flag index collection
The controller needs to bear physical and mental stress for completing the control task, the stress can be converted into time consumption, the borne stress is relieved through the time consumption, and the requirement for completing the objective task is met, and the time consumption is the size of the workload of the controller. In the measurable controller work time consumption, the land-air traffic channel occupancy rate is a flag index reflecting the controller work load.
The land-air communication channel occupancy rate refers to the land-air communication time length ratio in the unit time of the control sector. Setting the land-air communication of the control sector m times in unit time T, wherein the time length of the ith land-air communication is TiThe land-air communication channel occupancy rate is TrateThen, thenThe system is connected with the control voice data in a leading mode, the start time and the end time of communication between a controller and a pilot of a corresponding sector control seat are analyzed, then the duration of each communication is accumulated, the duration of the sector air-ground communication is obtained, and the occupancy rate of the air-ground communication channel is calculated.
Air traffic flow situation index prediction for control sector
The system synthesizes flight path data through a real-time leading air traffic control automation system and telegraph data in a fixed format of civil aviation flight dynamic state of a rebroadcasting system, obtains flight plan data of the aircraft, including flight number, takeoff time, takeoff airport, landing time, landing airport and other information, obtains the information of the takeoff flight in the air, such as height, speed, position and the like through the comprehensive flight path, and calculates the position information of the aircraft in the future period through a 4D flight path prediction technology. And obtaining data of future traffic flow situation indexes by calculating aircraft position information in a future period. In order to realize accurate prediction based on the 4D track prediction technology, the system establishes an aircraft basic information and operation performance database and an airway route information database.
Example four
The embodiment mode discloses a controller workload prediction method, which can be realized by selecting a hardware platform in the second embodiment, the selection of related workload prediction indexes is referred, and the collection of controller workload prediction sample data and control sector air traffic flow situation index prediction data can refer to the third embodiment.
In the embodiment, multiple linear regression models and multiple nonlinear regression models are simultaneously adopted, and the model with the minimum detection error is selected from the multiple linear regression models and the multiple nonlinear regression models to serve as the final prediction model.
(1) Multiple linear regression, which uses a linear function to fit multiple independent variables Xi(i-1, 2, …, n) and a single dependent variable Y, thereby determining a parameter b of the multiple linear regression modeli(i ═ 0,1,2, …, n), regressed into the original hypothesis equation, and the trend of the dependent variable was detected by the regression equation. The general form of the multiple linear regression model is:
Y＝b0+b1X1+b2X2+…+biXi+…+bnXn+ mu (formula 4.1)
Where μ is the random error except for the effect of the n independent variables on the dependent variable Y, obeying a normal distribution.
Assuming that the statistical sample has m sets of statistics, the matrix form of the multiple linear regression model can be expressed as:
y ═ XB + U (formula 4.2)
Wherein,
(formula 4.3)
(2) Multiple non-linear regression, assuming a non-linear relationship between independent variables (predictors) and dependent variables (controller workloads), the multiple non-linear model can be generally expressed as:
Y＝f[(b1,b2,…,bk)；X1,X2,…,Xn](formula 4.4)
The nonlinear regression function can adopt forms of quadratic function, power function, exponential function, hyperbolic function and the like according to the characteristics of sample data. The present embodiment is illustrated by a quadratic function:
Parameter b of multiple regression modeliAfter the estimation is performed, that is, after the sample regression function is solved, the sample regression function needs to be further subjected to statistical tests including fitting degree test, significance test, parameter confidence interval estimation and the like, then the detection error is calculated, and finally a model with a small error is selected as a final prediction model.
From the return of the coefficients of the decision R of each model2Value, F test and t test, respectively verifying and comparing the fitting degree and the significance of the two regression models, and calculating the detection degree and the significance of the two regression models on the basis of higher model fitting degree and obvious significanceAnd measuring errors, and selecting a model with the minimum error as a multiple regression model for the comprehensive detection of the operation performance of the sector.
The multiple regression-based controller workload prediction algorithm mainly comprises four parts, namely the construction of a regression model, the comparison and selection of the regression model, the prediction of the workload of a controller and the response alarm of the workload of the controller. Referring to fig. 9, the specific algorithm steps are:
step 1: selecting variables
Referring to the third embodiment, the input values of the 15 indexes are obtained according to the M groups of sample input data with the time duration of hour. Meanwhile, the occupancy rate of the land-air communication channel of the controller is used as the workload index Y of the controller. Examples of the obtained sample index data are as follows:
TABLE 1 example of controller workload prediction index sample data
Wherein the independent variable X ═ { X ═ XiAnd i is 1,2, … and 15, and the total number of the indexes is 15. X in Table 11～X15The 1-M group data and the M group data of Y are respectively substituted into formulas 4.3 and 4.4.
Wherein the sector trafficability detection index is { X }1,X2,X3,X4Respectively representing sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; sector complexity detection index is { X }5,X6,X7,X8Respectively representing the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed changing times of the sector aircraft and the navigation changing times of the sector aircraft; sector security detection index is { X9,X10Respectively representing a sector short-term conflict alarm frequency and a sector minimum safe height alarm frequency; sector economy detection index is { X11,X12,X13,X14,X15And the sector saturation, the sector queuing length, the sector aircraft delay frame rate, the sector aircraft delay time and the sector aircraft average delay time are respectively represented.
Step 2: data processing
Considering that different indexes have different dimensions and different magnitude order, in order to facilitate regression analysis of the model, standard conversion needs to be carried out on index data.
Let xij、x′ijRespectively representing the raw data of the ith sample and the normalized and converted data,sjrespectively, the mean and variance of the j (j ═ 1,2, …,15) th index data, then:
(formula 4.6)
Normalizing the data x 'after conversion'ijAs input data for regression analysis.
And step 3: constructing a regression model
Referring to the second and third embodiments, a multiple linear regression model and a multiple nonlinear regression model are respectively constructed, wherein the nonlinear regression model is in the form of a quadratic function. Obtaining parameter estimation values of two types of functions by fitting sample dataAnd sample output valueWhereinB represents formula 4.40～bnOr b of formula 4.50～b2nAn estimate of (d).
And 4, step 4: test regression model
From the return of the coefficients of the decision R of each model2And (3) value, F test and t test, respectively verifying and comparing the fitting degree and the significance of the two regression models, calculating the detection errors of the two regression models on the basis of higher model fitting degree and obvious significance, and selecting the model with the minimum error as the multiple regression model for the workload prediction of the controllers.
And 5: regression model result output
Obtaining the mean value of 15 indexes according to N groups of sample dataVariance sjPredicting data of air traffic flow situation indexes of control sectorsAs input data, and is subjected to a standardized conversion,is the processed input data. After the normalization process, willAnd importing the data into a controller workload prediction regression model to obtain a prediction result of the controller workload index.
Step 6: controller workload response alerts
According to the prediction result of the controller workload, responding to the alarm standard by referring to the set controller workload, and generating an alarm for the system when the alarm standard is met.
According to the algorithm process, 400 groups of all ACC01 sector related index data are collected, linear functions and nonlinear functions (quadratic functions) are respectively adopted to fit sample data, and R of the two functions is obtained through fitting calculation2P-value, and mean error, maximum error, minimum error, etc. of the fitting performance data. And respectively comparing the fitting effects of the controller workload prediction models based on the multiple linear regression model and the nonlinear regression model. The conclusion is as follows:
TABLE 2 multiple regression fitting effect comparison
From the above table, the indexes such as the fitting degree, the significance, and the error of the nonlinear function of the present embodiment are slightly better than those of the linear function. Therefore, a non-linear function is chosen here as a predictive model of the controller workload. The fitting effect graph and the prediction error graph of the model are shown in FIGS. 10 and 11:
in summary, the controller workload prediction model based on multiple nonlinear regression is:
(formula 4.7)
According to equation 4.7, the administrator workload is predicted. And acquiring traffic flow situation index data of 5 time periods in the future according to the air traffic flow situation prediction result of the control sector. After normalization, which is carried out in equation 4.7, the forecasted controller workload for the next 5 time slots is calculated as shown in the table below.
TABLE 3 analysis of example forecasts of controller workload
And according to the alarm standard of the workload of the controller, carrying out corresponding alarm when the workload of the controller reaches the alarm standard in the future time period.
After the prediction method and the corresponding system are put into operation, corresponding management is required. The recommended system management is as follows:
managing user authority, distributing user name and authority to each user, ensuring data safety and preventing data leakage.
Each user corresponds to 0 to a plurality of roles, and each role can be flexibly allocated with access and operation authorities by management personnel.
Setting parameters necessary for system operation, including map parameters, telegraph processing and radar data processing parameters, long-term schedule, system display parameter setting and other parameters required to be set.
Providing log management function to record system operation and keep important data operation information. The method comprises the following steps: the device comprises a log recording module, a log query module and a log backup and clearing module.
Providing parameter configuration function, and providing tool for system maintenance personnel.
Sixthly, providing a data import and export function.
The proposed configuration for implementing the prediction method and system of the present invention is as follows:
EXAMPLE five
The controller workload prediction method of the embodiment comprises the following steps:
and constructing a multiple regression model according to the relevant indexes of the controller workload prediction and the sample data of the controller workload prediction. Obtaining input values of relevant indexes of the workload prediction of the controllers according to M groups of sample input data with the duration of hours; and taking the occupancy rate of the land-air communication channel of the controller as the workload index of the controller to obtain the workload prediction sample data of the controller.
And importing the air traffic flow situation index prediction data of the control sector into a multiple regression model.
Considering that different dimensions and magnitude difference exist among different indexes, before constructing a multiple regression model, firstly, carrying out standardized conversion on input relevant indexes of controller workload prediction and controller workload prediction sample data; correspondingly, standardized conversion is carried out on the forecast data of the air traffic flow situation indexes of the control sector introduced with the multiple regression model; this may facilitate regression analysis of the model.
The multiple regression model comprises a multiple linear regression model and a multiple non-linear regression model, wherein the non-linear regression model adopts a quadratic function. The model function is referred to the above embodiments.
Fitting the workload prediction sample data of a controller through a multiple linear regression model and a multiple nonlinear regression model respectively to obtain two groups of sample regression functions, and performing statistical test on the two groups of sample regression functions, wherein the statistical test step comprises the fitting degree and significance test, and when the fitting degree and the significance exceed preset values, the detection error is calculated.
And importing the air traffic flow situation index prediction data of the control sector into a multiple regression model with the minimum detection error to obtain a prediction result.
When the prediction result exceeds the threshold value, the controller workload responds to the alarm.
The control sector air traffic flow situation indexes comprise sector operation trafficability indexes, sector operation complexity indexes, sector operation safety indexes and sector operation economy indexes; the sector trafficability detection indexes respectively comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; the sector complexity detection indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft; the sector safety detection indexes comprise sector short-term conflict alarm frequency and sector minimum safety height alarm frequency; the sector economy detection indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time.
As shown in fig. 12, the present embodiment also discloses a system for predicting the workload of the administrator. The system comprises an operation index detection database, a data leading device and an index acquisition device, wherein the data leading device is coupled with the operation index detection database of the control sector.
The data leading device comprises a telegraph data interface, a comprehensive track data interface and a control voice data interface which are respectively coupled with the control sector operation index detection database; the index acquisition device is used for acquiring the air traffic flow situation indexes of the control sector, and the air traffic flow situation indexes of the control sector comprise a sector operation trafficability index, a sector operation complexity index, a sector operation safety index and a sector operation economy index.
A component module: selecting air traffic flow situation indexes of a control sector, substituting controller workload prediction sample data corresponding to relevant indexes into a multiple linear regression model and a multiple nonlinear regression model for fitting; obtaining parameter estimation values of a multiple linear regression model and a multiple nonlinear regression model; calculating a detection error through statistical test, and determining a multiple regression model for predicting the workload of the controller;
a prediction module: and importing the air traffic flow situation index prediction data of the control sector into a controller workload prediction multiple regression model to obtain a prediction result of the controller workload.
A standardization conversion module: the system is used for carrying out standardized conversion on the sample data and the real-time data of the air traffic flow situation indexes of the control sector;
an alarm module: when the prediction result exceeds the threshold value, the controller workload responds to the alarm.
The component module reads related indexes of controller workload prediction and sample data of controller workload prediction from a control sector operation index detection database; and the prediction module reads the prediction data of the air traffic flow situation indexes of the control sector from the control sector operation index detection database.
The index acquisition device is used for acquiring air traffic flow situation indexes of a control sector, wherein the air traffic flow situation indexes of the control sector comprise a sector operation trafficability index, a sector operation complexity index, a sector operation safety index and a sector operation economy index; the sector trafficability detection indexes respectively comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; the sector complexity detection indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft; the sector safety detection indexes comprise sector short-term conflict alarm frequency and sector minimum safety height alarm frequency; the sector economy detection indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.
Claims (11)
1. A method for administrator workload prediction, comprising the steps of:
step 1: selecting air traffic flow situation indexes of a control sector at a certain time interval and controller workload indexes corresponding to the situation indexes as sample data;
step 2: establishing a linear regression model and a nonlinear regression model according to the sample data;
and step 3: comparing the linear regression model with the nonlinear regression model through the fitting degree, the significance and the error analysis to determine a multiple regression model for predicting the workload of the controller;
and 4, step 4: and importing the real-time control sector air traffic flow situation indexes into a controller workload prediction multiple regression model to obtain controller workload indexes.
2. The controller workload prediction method according to claim 1, characterized in that the controlled sector air traffic flow situation indicators in step 1 comprise a sector operation trafficability indicator, a sector operation complexity indicator, a sector operation safety indicator and a sector operation economy indicator.
3. The controller workload prediction method according to claim 2, wherein the control sector air traffic flow situation indicators comprise a sector operation traffic performance indicator, a sector operation complexity indicator, a sector operation safety indicator, and a sector operation economy indicator;
the sector trafficability detection indexes respectively comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density;
the sector complexity detection indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft;
the sector safety detection indexes comprise sector short-term conflict alarm frequency and sector minimum safety height alarm frequency;
the sector economy detection indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time.
4. A method for forecasting the workload of a controller as claimed in claim 1 or 3, characterized in that in step 2, the sample data is subjected to a standardized transformation; the normalization conversion process is as follows:
let xij、x′ijRespectively representing the raw data and normalized converted number of the ith sampleAccording to the above-mentioned technical scheme,sjrespectively representing the mean and variance of the jth index data, then:
5. the controller workload prediction method according to claim 1, wherein the step 2 specifically comprises:
according to the standardized sample data x'ij(i 1,2, … m, j 1,2, … n), establishing a multiple linear regression model and a multiple non-linear regression model respectively, and solving a coefficient bi，
Wherein the multiple linear regression model is:
y ═ XB + U (formula 1)
Wherein,
the multiple nonlinear regression model is:
Y＝f[(b1,b2,…,bk)；X1,X2,…,Xn](formula 2)
The method comprises the steps that a dependent variable Y is a control sector operation performance comprehensive index, an independent variable X is n control sector operation performance comprehensive detection indexes, m represents a control sector operation performance index sample under m groups of time intervals, U is a random error except for the influence of m independent variables on the dependent variable Y and obeys normal distribution, and f represents a nonlinear regression function.
6. The controller workload prediction method according to claim 1, wherein the step 3 specifically comprises:
from the return of the coefficients of the decision R of each model2Value, F test,And t, respectively verifying and comparing the fitting degree and the significance of the two regression models, calculating the detection errors of the two regression models on the basis of higher model fitting degree and obvious significance, and selecting the model with the minimum error as the multiple regression model for the workload prediction of the controllers.
7. The controller workload prediction method according to claim 1, characterized in that the real-time control sector air traffic flow situation indicators in step 4 are subjected to a standardized transformation before being input into the multiple regression model; the normalization conversion process is as follows:
mean value of n indexes of sample data according to m groups of time intervalsVariance sjFor the air traffic flow situation index t of the control sectorj(j ═ 1, 2.. times, n) normalized conversion:converting the data tj' import into controller workload prediction multiple regression model.
8. The controller workload prediction method according to claim 1, characterized in that the method further comprises a step 5 of the controller workload responding to an alarm when the controller workload index exceeds a threshold value.
9. A system for forecasting workload of an administrator, comprising,
a component module: selecting air traffic flow situation indexes of a control sector, substituting controller workload prediction sample data corresponding to relevant indexes into a multiple linear regression model and a multiple nonlinear regression model for fitting; obtaining parameter estimation values of a multiple linear regression model and a multiple nonlinear regression model; calculating a detection error through statistical test, and determining a multiple regression model for predicting the workload of the controller;
a prediction module: and importing the air traffic flow situation index prediction data of the control sector into a controller workload prediction multiple regression model to obtain a prediction result of the controller workload.
10. The system for forecasting controller workload as claimed in claim 9, further comprising a normalization conversion module: the system is used for carrying out standardized conversion on the sample data and the real-time data of the air traffic flow situation indexes of the control sector;
an alarm module: when the prediction result exceeds the threshold value, the controller workload responds to the alarm.
11. The system for forecasting a traffic controller workload according to claim 10, further comprising a control sector traffic flow situation detection database, a data lead-in device and an index collection device coupled to said control sector traffic flow situation detection database;
the data leading device comprises a telegraph data interface, a comprehensive track data interface and a control voice data interface which are respectively coupled with the control sector traffic flow situation detection database;
the index acquisition device is used for acquiring air traffic flow situation indexes of a control sector, wherein the air traffic flow situation indexes of the control sector comprise a sector operation trafficability index, a sector operation complexity index, a sector operation safety index and a sector operation economy index; the sector trafficability detection indexes respectively comprise sector flow, sector navigation mileage, sector navigation time and sector traffic flow density; the sector complexity detection indexes comprise the climbing times of the sector aircraft, the descending times of the sector aircraft, the speed change times of the sector aircraft and the navigation change times of the sector aircraft; the sector safety detection indexes comprise sector short-term conflict alarm frequency and sector minimum safety height alarm frequency; the sector economy detection indexes comprise sector saturation, sector queuing length, sector aircraft delay frame rate, sector aircraft delay time and sector aircraft average delay time;
the component module reads the related indexes of the controller workload prediction and the sample data of the controller workload prediction from the traffic flow situation detection database of the control sector; the prediction module reads the air traffic flow situation index prediction data of the control sector from the traffic flow situation detection database of the control sector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510645215.3A CN105205565A (en) | 2015-09-30 | 2015-09-30 | Controller workload prediction method and system based on multiple regression model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510645215.3A CN105205565A (en) | 2015-09-30 | 2015-09-30 | Controller workload prediction method and system based on multiple regression model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105205565A true CN105205565A (en) | 2015-12-30 |
Family
ID=54953234
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510645215.3A Pending CN105205565A (en) | 2015-09-30 | 2015-09-30 | Controller workload prediction method and system based on multiple regression model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105205565A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107944472A (en) * | 2017-11-03 | 2018-04-20 | 北京航空航天大学 | A kind of airspace operation situation computational methods based on transfer learning |
| CN108171379A (en) * | 2017-12-28 | 2018-06-15 | 无锡英臻科技有限公司 | A kind of electro-load forecast method |
| CN111091244A (en) * | 2019-12-16 | 2020-05-01 | 武汉材料保护研究所有限公司 | Engine lubricating oil change period prediction method |
| CN111126685A (en) * | 2019-12-16 | 2020-05-08 | 武汉材料保护研究所有限公司 | Method for establishing engine lubricating oil quality prediction model |
| CN111968414A (en) * | 2020-08-26 | 2020-11-20 | 成都民航空管科技发展有限公司 | 4D trajectory prediction method and device based on big data and AI and electronic equipment |
| CN114530059A (en) * | 2022-01-14 | 2022-05-24 | 南京航空航天大学 | Dynamic configuration method and system for multi-sector monitoring seat |
| CN117978916A (en) * | 2024-04-01 | 2024-05-03 | 中国民用航空飞行学院 | Method and equipment for predicting call load of controller |
-
2015
- 2015-09-30 CN CN201510645215.3A patent/CN105205565A/en active Pending
Non-Patent Citations (6)
| Title |
|---|
| 何元清 等: "《大学计算机基础》", 31 August 2013 * |
| 周品: "《MATLAB概率与数理统计》", 30 November 2012 * |
| 李鸿吉: "《Visual Basic 6.0数理统计实用算法》", 30 September 2003, 科学出版社 * |
| 袁霄 等: ""基于扇区动态复杂性因素的航空管制员工作负荷计算"", 《安全与环境学报》 * |
| 郭少英 等: "《市场调研与分析技术》", 31 October 2008 * |
| 麦好: "《机器学习实践指南 案例应用解析》", 30 April 2014 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107944472B (en) * | 2017-11-03 | 2019-05-28 | 北京航空航天大学 | A kind of airspace operation situation calculation method based on transfer learning |
| CN107944472A (en) * | 2017-11-03 | 2018-04-20 | 北京航空航天大学 | A kind of airspace operation situation computational methods based on transfer learning |
| CN108171379B (en) * | 2017-12-28 | 2021-12-17 | 无锡英臻科技有限公司 | Power load prediction method |
| CN108171379A (en) * | 2017-12-28 | 2018-06-15 | 无锡英臻科技有限公司 | A kind of electro-load forecast method |
| CN111091244A (en) * | 2019-12-16 | 2020-05-01 | 武汉材料保护研究所有限公司 | Engine lubricating oil change period prediction method |
| CN111126685A (en) * | 2019-12-16 | 2020-05-08 | 武汉材料保护研究所有限公司 | Method for establishing engine lubricating oil quality prediction model |
| CN111091244B (en) * | 2019-12-16 | 2023-02-03 | 武汉材料保护研究所有限公司 | Engine lubricating oil change period prediction method |
| CN111126685B (en) * | 2019-12-16 | 2023-02-03 | 武汉材料保护研究所有限公司 | Method for establishing engine lubricating oil quality prediction model |
| CN111968414A (en) * | 2020-08-26 | 2020-11-20 | 成都民航空管科技发展有限公司 | 4D trajectory prediction method and device based on big data and AI and electronic equipment |
| CN111968414B (en) * | 2020-08-26 | 2022-08-05 | 成都民航空管科技发展有限公司 | 4D trajectory prediction method and device based on big data and AI and electronic equipment |
| CN114530059A (en) * | 2022-01-14 | 2022-05-24 | 南京航空航天大学 | Dynamic configuration method and system for multi-sector monitoring seat |
| CN114530059B (en) * | 2022-01-14 | 2023-03-10 | 南京航空航天大学 | Dynamic configuration method and system for multi-sector monitoring seat |
| CN117978916A (en) * | 2024-04-01 | 2024-05-03 | 中国民用航空飞行学院 | Method and equipment for predicting call load of controller |
| CN117978916B (en) * | 2024-04-01 | 2024-05-28 | 中国民用航空飞行学院 | Method and equipment for predicting call load of controller |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105205565A (en) | Controller workload prediction method and system based on multiple regression model | |
| CN105225539B (en) | The method and system of sector runnability composite index based on principal component analysis | |
| CN105160201A (en) | Genetic algorithm back propagation (GABP) neural network based controller workload prediction method and system | |
| CN105261240B (en) | A kind of sector runnability method for comprehensive detection and system based on cluster analysis | |
| CN106156913A (en) | Health control method for aircraft department enclosure | |
| CN105225007A (en) | A kind of sector runnability method for comprehensive detection based on GABP neural network and system | |
| CN106023655A (en) | Sector air traffic congestion state monitoring method | |
| CN104835354A (en) | Control load management system and controller workload evaluation method | |
| CN103903101A (en) | General aviation multi-source information supervisory platform and method | |
| CN110060513A (en) | Workload for air traffic controllers appraisal procedure based on historical trajectory data | |
| CN105225193A (en) | A kind of method and system of the sector runnability aggregative index based on multiple regression model | |
| CN105118333B (en) | A kind of air traffic control analog simulation method for detecting abnormality and device based on multiple regression model | |
| CN105206115B (en) | A kind of air traffic control sector traffic capacity detection method based on principal component analysis | |
| Wieland et al. | Predicting sector complexity using machine learning | |
| Mott et al. | Estimation of aircraft operations at airports using nontraditional statistical approaches | |
| Wanke et al. | Modeling traffic prediction uncertainty for traffic management decision support | |
| Torres | Determination and ranking of trajectory accuracy factors | |
| CN110928752A (en) | Method, device and equipment for evaluating health degree of air traffic control station | |
| Borjalilu et al. | Entropy-based model for aerodromes safety risk assessment to implement safety management systems | |
| CN116258381A (en) | Quantitative evaluation method and device for operation command work | |
| Paglione et al. | Lateral intent error’s impact on aircraft prediction | |
| Valdés et al. | Bayesian networks for decision-making and causal analysis under uncertainty in aviation | |
| Temme et al. | Factors for Pilot's Decision Making Process to Avoid Severe Weather during Enroute and Approach | |
| Manning et al. | Relationship of sector activity and sector complexity to air traffic controller taskload | |
| Šimić et al. | Airport delay prediction using machine learning regression models as a tool for decision making process |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151230 |
|
| RJ01 | Rejection of invention patent application after publication |