A quantitative evaluation method and system for controller status in less than radar interval event investigation

Abstract

The present invention discloses a method and system for quantitatively evaluating the state of a controller in an investigation of an event less than a specified radar interval. The evaluation method includes extracting causal keywords, constructing an improved HFACS model based on the HFACS model and the causal keywords, and constructing a controller state evaluation index system based on the improved HFACS model; determining the weight of the controller state evaluation index; using the Z-Score normalization method to calculate the danger index of the controller state evaluation index, and formulating an evaluation standard for the controller state. The present invention comprehensively analyzes the multiple factors that affect the controller state index, constructs a scientific and objective quantitative evaluation tool for the controller state danger index for the investigation of events less than the specified radar interval, deeply reveals the intertwined relationship between human factors and other factors, provides a more accurate controller state evaluation system, and provides a solid basis for aviation safety management decisions, which can effectively prevent the recurrence of similar incidents and comprehensively improve aviation safety standards.

Description

Technical Field

[0001] The present invention relates to the technical field of civil aviation safety management and event investigation, and in particular to a method and system for quantitatively evaluating the state of a controller in an event investigation less than a radar interval. Background Art

[0002] In civil aviation control, radar separation refers to the minimum safe distance between aircraft within the radar coverage area. An event of less than specified radar separation means that the distance between two aircraft is less than the minimum specified value, which may lead to flight safety hazards. Therefore, the safety of radar separation is extremely important.

[0003] At present, the occurrence of events less than the prescribed radar interval is often due to the failure of key aircraft parameters to meet standards and poor controller status. The existing hazard index assessment method has problems such as too few indicators and lack of scientificity in terms of controller status. It does not fully consider the human factors in the occurrence and development of events, as well as the logical influence relationship between human factors and environmental factors, equipment factors, and management factors. The assessment of the controller status hazard index is often based on subjective judgment based on experience, and it is impossible to systematically and quantitatively analyze and evaluate the controller's status when an event less than the prescribed radar interval occurs, making it difficult to improve the level of aviation safety management in a high-quality manner. Summary of the invention

[0004] The present invention aims at the technical problem in the prior art that it is difficult to quantitatively and systematically analyze and evaluate the state of a controller for an event less than a radar separation, and provides a method and system for quantitatively evaluating the state of a controller for an event less than a radar separation.

[0005] In view of the above technical problems, an embodiment of the present invention provides a method for quantitatively evaluating the state of a controller for investigating an event less than a radar interval, comprising the following steps:

[0006] Extract causal keywords based on the investigation report of less than radar interval incidents;

[0007] An improved HFACS model is constructed based on the HFACS model and causal keywords, and a controller status evaluation index system is constructed based on the improved HFACS model;

[0008] Based on the controller status evaluation index system, determine the weights of the controller status evaluation index;

[0009] According to the weight of the controller status evaluation index, the Z-Score normalization method is used to calculate the danger index of the controller status evaluation index;

[0010] Based on the danger index of the controller status evaluation indicators, the evaluation criteria for the controller status are formulated.

[0011] The present invention also provides a controller status quantitative assessment system for less than radar interval event investigation, comprising:

[0012] An extraction module, used to extract causal keywords based on the investigation report of the less than radar interval incident;

[0013] Improved HFACS model construction module, used to construct an improved HFACS model based on the HFACS model and causal keywords, and to construct a controller status evaluation index system based on the improved HFACS model;

[0014] A weight determination module, used to determine the weight of the controller status evaluation index based on the controller status evaluation index system;

[0015] A danger index calculation module is used to calculate the danger index of the controller status evaluation index using a Z-Score normalization method according to the weight of the controller status evaluation index;

[0016] The evaluation module is used to formulate the evaluation criteria of the controller status based on the danger index of the controller status evaluation index.

[0017] In the present invention, causal keywords are extracted through text mining technology to accurately identify the main causal factors of events less than the radar interval, and potential causal factors that are not easily detected are mined through analysis of a large amount of text data to improve the accuracy of the assessment, which can deeply reveal the complex relationship between human factors and other factors; by introducing the HFACS model and combining it with causal keywords for improvement, an improved HFACS model is obtained to improve the accuracy and objectivity of the assessment, so as to enhance the pertinence and relevance of the controller status assessment index; the danger index of the controller status assessment index is determined by the weight of the controller status assessment index, and by quantitatively analyzing multiple indicators that affect the controller status, the indicator weights are scientifically calculated in combination with complex network theory, which overcomes the previous experience-based subjective judgment method, making the calculation result of the danger index more objective and accurate, and then scientifically and reasonably dividing the event level according to the danger index, providing decision support for safety management and accident prevention. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Schematic diagram of a flow chart of a method for quantitatively evaluating a controller's status in one embodiment of the present invention;

[0019] Figure 2 is a structural diagram of a controller status quantitative evaluation system in one embodiment of the present invention;

[0020] Figure 3 is a schematic diagram of the marking of the controller status evaluation index in one embodiment of the present invention;

[0021] Figure 4The figure is a schematic diagram of the network topology structure of the controller status evaluation index in one embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] In one embodiment, if Figure 1 As shown, the present invention provides a method for quantitatively evaluating the state of a controller for investigating an event less than a radar interval, comprising the following steps S10-S50:

[0024] S10, extracting causal keywords based on the investigation report of the less than radar interval event; mining and analyzing the investigation report of the less than radar interval event, obtaining keywords related to causal factors and extracting causal keywords therefrom;

[0025] S20. Construct an improved HFACS model based on the HFACS model and causal keywords, and construct a controller status evaluation index system based on the improved HFACS model;

[0026] S30, determining the weight of the controller status evaluation index based on the controller status evaluation index system;

[0027] S40, according to the weight of the controller status evaluation index, the danger index of the controller status evaluation index is calculated by using the Z-Score normalization method;

[0028] S50. Based on the hazard index of the controller status evaluation index, formulate the evaluation criteria for the controller status.

[0029] In the present invention, causal keywords are extracted through text mining technology to accurately identify the main causal factors of events less than the radar interval, and potential causal factors that are not easily detected are mined through analysis of a large amount of text data to improve the accuracy of evaluation, which can deeply reveal the complex relationship between human factors and other factors; by introducing the HFACS model and combining it with causal keywords for improvement, an improved HFACS model is obtained to improve the accuracy and objectivity of the evaluation, so as to enhance the pertinence and relevance of the controller status evaluation index; the danger index of the controller status evaluation index is determined by the weight of the controller status evaluation index, and by quantitatively analyzing multiple indicators that affect the controller status, the indicator weight is scientifically calculated in combination with complex network theory, which overcomes the previous experience-based subjective judgment method, making the calculation result of the danger index more objective and accurate, and then scientifically and reasonably dividing the event level according to the danger index, providing decision support for safety management and accident prevention.

[0030] In one embodiment, the step S10 includes the following steps S101-S103:

[0031] S101. Preprocessing the investigation report of the less than radar interval event, wherein the preprocessing includes cleaning the text of the investigation report and removing stop words, punctuation marks and irrelevant characters. It is understandable that the above preprocessing can significantly reduce the noise of the text data and improve the efficiency and accuracy of subsequent text processing.

[0032] S102. Segmenting the text into words or phrases helps to refine the granularity of text analysis and make keyword extraction more accurate; removing common and meaningless words and creating or obtaining a stop word list. Removing these words can highlight the core content of the text and improve the pertinence of keyword extraction; understandably, a comprehensive and accurate stop word list can more effectively reduce text noise.

[0033] S103, after filtering the stop words in the stop word list in the text, use the non-stop words adjacent to the text as candidate keywords, calculate the keyword score based on the keyword frequency and co-occurrence frequency, sort all candidate keywords according to the score, and select the keyword with the highest score as the final causal keyword of the controller status assessment index of the less than radar interval event. It can be understood that the calculation of the keyword score based on the keyword frequency and co-occurrence frequency, and sorting all candidate keywords according to the score, includes the following steps:

[0034] The frequency of occurrence of keywords F is calculated through the word frequency calculation model c , the calculation expression of the word frequency calculation model is:

[0035]

[0036] Among them, Fc represents the number of times word i appears in document j, Represents the sum of the occurrences of all causal keywords in document j.

[0037] Calculate the co-occurrence frequency GF of two causal keywords m , used to evaluate the degree of association between two words, mainly by calculating the co-occurrence frequency of two adjacent non-stop words, GF i Represents the average co-occurrence frequency of word i and its adjacent non-stop words;

[0038]

[0039] Wherein, k represents the number of pairs of word i and its adjacent non-stop words, m=1,...k.

[0040] Based on word frequency F c and the co-occurrence frequency GF m , and the score of the causal keyword is calculated through the scoring model, and the calculation expression of the scoring model is:

[0041] Score i =F c α+GF i β

[0042] Among them, α and β are the weights of word frequency and co-occurrence frequency respectively, Score i Indicates the final score of word i. According to the above scores, the scores are arranged from small to large.

[0043] Understandably, by calculating the keyword score based on the keyword frequency and co-occurrence frequency, the causal keywords that are highly related to the less than radar interval event can be accurately identified. The word frequency reflects the frequency of a causal keyword in the text, and the co-occurrence frequency reveals the closeness of the co-occurrence of different causal keywords in the same text. By combining the two, the truly representative and critical causal keywords can be accurately extracted. The results of the causal keyword score and ranking provide an important quantitative basis for the assessment of the controller's status. By analyzing the causal keywords, we can gain an in-depth understanding of the human factors, environmental factors, management factors and their mutual influence that lead to the less than radar interval event, and then make a more comprehensive and accurate assessment of the controller's status in the event.

[0044] In one embodiment, the step S20 includes the following steps S201-S203:

[0045] S201, replace the specific indicators of each level of the HFACS model with the causal keywords, and replace the inadequate supervision, inappropriate operation plan, no correction of problems and supervision violations in the unsafe supervision indicators in the HFACS model with unreasonable team matching, fatigue / overtime work, lack of on-site management, regular spot checks, untimely policy dissemination and incomplete skills training; understandably, by using the causal keywords based on the event investigation report, the improved HFACS model is more matched with the specific causes of the less than radar interval event, and can more accurately reflect the behavior and potential influencing factors of the controller in such events. The replaced indicators are directly related to the actual management problems and personnel status problems that lead to the less than radar interval event, which helps to identify the key causes and formulate corresponding prevention and management measures. The improved HFACS model is adjusted based on actual event data, making the evaluation process more systematic and quantitative, and improving the scientificity and rationality of the model. In this way, using the improved HFACS model for event investigation can locate the specific causes of the less than radar interval event more quickly and reduce errors caused by subjective judgment.

[0046] S202, adding an unsafe behavior performance indicator layer to the unsafe behavior indicator layer of the HFACS model;

[0047] The unsafe behavior indicator layer includes improper information processing, improper decision-making and improper action execution; the unsafe behavior performance indicator layer includes failure to effectively grasp environmental information, failure to effectively grasp operation dynamics, unreasonable control plan, failure to adjust command for operation conflict, issuance of wrong control instructions, failure to effectively monitor repetition and failure to ensure consistency between movement status and process order, so as to constitute an improved HFACS model; understandably, after fully considering the preconditions of unsafe behavior and the occurrence and development of unsafe behavior events, adding an unsafe behavior performance indicator layer can enhance the meticulousness and accuracy of the analysis and significantly improve it. In the previous unsafe behavior indicator system, there was a major problem of classification confusion. Originally, unsafe behaviors were roughly divided into two categories: errors and violations. For example, errors can be further subdivided into decision errors, skill errors and perceptual errors. However, in actual event analysis, the occurrence of an event often involves multiple error types at the same time, which makes the classification vague and difficult to define. For example, an event may include errors in decision-making, reflect deficiencies in skills, or even errors in perceptual judgment, so the previous classification method not only increases the complexity of statistics, but also reduces the accuracy of analysis.

[0048] In addition, there are also problems with the classification of violations. The distinction between habitual violations and accidental violations is too subjective and lacks objective standards and basis. This subjectivity greatly affects the fairness and effectiveness of incident investigations.

[0049] Therefore, we chose to thoroughly improve the unsafe behavior indicator system. Specifically, we will add an unsafe behavior performance indicator layer to construct the performance of specific events. Through this improvement, we can more clearly explain the occurrence, development and performance process of the event. This indicator system based on event performance not only simplifies the classification logic, but also improves the efficiency and accuracy of event investigation; more importantly, it helps us to better analyze human factors, thereby providing strong support for preventing similar incidents from happening again. S203. The correlation between the controller status evaluation indicators at each level of the improved HFACS model that replaces the causal keywords and adds a safe behavior performance indicator layer is sorted out to obtain the causal chain of events less than the radar interval, and to construct a controller status evaluation indicator system. It can be understood that the improved HFACS model has not only been modified in terms of the level and comprehensiveness of the indicators, but also the relationship between indicators at different levels has been sorted out, so that the causal chain of the event is clearly visible, such as Figure 3 As shown, insufficient resources required for safety management → fatigue / overtime work → poor physical and mental state of controllers → failure to effectively grasp dynamic operation information → improper information preprocessing → unsafe behavior, a total of 465 event cause chains were obtained.

[0050] In one embodiment, the step S30 includes the following steps S301-S303:

[0051] S301, constructing a complex network based on the causal chain in the controller status evaluation index system, and performing topological feature analysis on the complex network; the complex network includes nodes and edges, the nodes represent the controller status evaluation indicators of each level of the improved HFACS model, and the edges represent the connection relationship between the nodes; Figure 4 As shown in the figure, by sorting out the relationship between the indicators in the causal chain of 465 events in the controller status evaluation indicator system, a small complex network with 28 nodes and 82 edges is constructed. The nodes correspond to specific indicators in each level, and the edges represent the relationship or mutual influence between the nodes.

[0052] In one embodiment, network scale analysis can understand the overall complexity and interaction range of each factor in the system, node degree analysis can identify key indicators in the network, and degree distribution can determine whether the network has scale-free characteristics (i.e., a few nodes have high connectivity, and most nodes have few connections); scale-free analysis can determine whether there are hub nodes in the network; and similar matching analysis can reflect the organizational structure characteristics of the network. Then, the importance of the nodes in the network is analyzed, including degree centrality (used to determine the degree of connection between the node and other nodes), betweenness centrality (used to determine whether it is a key bridge node in the network) and Pagerank value (used to determine the stability of the network); finally, the three types of indicators, degree centrality, betweenness centrality, and Pagerank value, are standardized and averaged to obtain the comprehensive value of the importance of the nodes in the complex network caused by the radar interval event, reflecting the global influence of the indicator in the entire network.

[0053] The step S301 includes the following steps S3011-S3014:

[0054] S3011, the topological feature analysis includes network model analysis, node degree analysis, scale-free analysis and similar matching analysis,

[0055] The network model analysis includes the following steps:

[0056] Define key metrics of the network model, including the average path length d avg and the network diameter d diameter , the average path length d avg It is used to measure the average distance between two nodes in a complex network. diameter Used to measure the distance between the two farthest nodes in a complex network; the average path length d avg and the network diameter d diameter The calculation expressions are:

[0057]

[0058] d diameter = max d ij (2)

[0059] Among them, d avg represents the average distance of all node pairs in the complex network, d ij represents the number of edges of the shortest path between node i and node j, n(n-1) represents the number of all node pairs in a directed network, n represents the number of nodes in a complex network, d diameter Represents the maximum distance among all pairs of nodes in a complex network;

[0060] Calculate the key indicators and use formula (1) and formula (2) to calculate the average path length d of all node pairs in the complex network avg and the network diameter d diameter ;

[0061] Analyze the compactness, scalability, and overall size of complex networks based on the calculated average path length and network diameter;

[0062] S3012, the node degree analysis comprises the following steps:

[0063] Define node degree. Node degree is the number of edges connected to the node in a complex network. For a directed network, the node degree includes the out-degree. and in-degree The out-degree It refers to the sum of the number of connected edges starting from the node. It refers to the sum of the number of connected edges ending at the node. The node degree is the sum of the out-degree and the in-degree. The calculation expressions of the node degrees are:

[0064]

[0065] Among them, l ij represents a directed edge from node i to node j, l ij represents the directed edge from node j to node i, and n represents the number of nodes in the complex network;

[0066] Calculate the node degree and construct the degree distribution function. The degree distribution p(k) refers to the proportion of nodes with degree k in the complex network in the entire complex network. The calculation expression of the degree distribution p(k) is:

[0067]

[0068] Where n represents the number of nodes in the complex network, n k represents the number of nodes with degree k;

[0069] The statistically obtained degree distribution data is plotted into a curve graph, the shape of the degree distribution curve is analyzed, and whether the complex network has power-law distribution characteristics is determined;

[0070] S3013, the scale-free analysis comprises the following steps:

[0071] The cumulative degree distribution is calculated by the cumulative degree distribution function, which represents the probability score of nodes with a degree value not less than k. The calculation expression of the cumulative degree distribution function is:

[0072]

[0073] Among them, p(i) represents the number of nodes with degree i;

[0074] Draw a cumulative degree distribution curve based on the calculated cumulative degree distribution data, analyze the characteristics of the network node degree distribution, and if the cumulative degree distribution function or the degree distribution function satisfies the power law distribution, then the complex network is determined to be a scale-free network;

[0075] The method for determining whether a complex network is a scale-free network includes: performing linear fitting on the degree distribution or cumulative degree distribution of the complex network in a double logarithmic coordinate system, and the calculation model of the linear fitting is:

[0076] P(k)~k -γ (6)

[0077] Where γ represents the power law exponent, and the value range of γ is between 2 and 3. If the data points are arranged along a straight line, it indicates that the cumulative degree distribution or degree distribution of the complex network follows a power law distribution.

[0078] S3014, the same type matching analysis includes the following steps:

[0079] Specify the properties of nodes in complex networks, including node degree, average neighbor degree or other specific characteristics;

[0080] Observe the connection patterns between nodes in complex networks, and observe the connection tendencies between high-degree nodes and low-degree nodes;

[0081] The same type matching coefficient r is calculated, and the coefficient r is used to quantify the degree of the same type matching in the network. The calculation expression of the same type matching coefficient r is:

[0082]

[0083] Among them, j i and k i Represents the degree of two nodes j and k connected by edge i, and the value of r ranges from -1 to 1;

[0084] If r>0, it indicates that the complex network exhibits positive homogeneity, that is, nodes with high similarity tend to be connected together;

[0085] If r < 0, it indicates that the complex network exhibits negative homogeneity, that is, nodes of different types tend to be connected together.

[0086] Understandably, by analyzing the scale of the network, we can intuitively grasp the overall complexity and scale of the system, laying the foundation for subsequent in-depth analysis. Node degree analysis and degree distribution further reveal the density and distribution of node connections in the network, which helps to understand the microstructure of the network. Scale-free analysis and similar matching analysis can accurately identify hub nodes and specific organizational structure characteristics in the network. Hub nodes usually occupy a core position in the network and have an important impact on the dissemination of information, the allocation of resources, etc.; while similar matching reveals the similarities or differences between nodes in the network, which is of great significance for understanding the formation mechanism and dynamic changes of the network.

[0087] S302, performing importance analysis on the nodes of the complex network according to the result of topological feature analysis of the complex network, wherein the importance analysis includes analysis of degree centrality for determining the degree of connection between a node and other nodes, betweenness centrality for determining whether it is a key bridge node in the complex network, and Pagerank value for determining the stability of the network;

[0088] In one embodiment, the step S302 includes the following steps S3021-S3023:

[0089] S3021. The degree centrality of the node is defined as the number of neighbor nodes directly connected to any node in the complex network. The normalized calculation formula of the degree centrality is:

[0090]

[0091] Where n represents the total number of nodes in the complex network, k i represents the degree of node i, and n-1 represents the number of other nodes connected to node i besides itself;

[0092] S3022: The betweenness centrality of the node is defined as the ratio of the number of all shortest paths in the complex network that pass through node i, and the calculation formula is:

[0093]

[0094] Among them, δ km is the number of shortest paths between node k and node m, δ km (i) is δ km The number of shortest paths passing through node i in ;

[0095] S3023. The node with a high Pagerank value is defined as a node pointed to by multiple other important nodes. The calculation formula of the Pagerank value is:

[0096]

[0097] Among them, PR(i) is the Pagerank value of node i, d=0.85, d is the damping coefficient, M(i) is the set of all nodes pointing to node i, and L(j) is the out-degree of node j.

[0098] S303. Standardize the degree centrality, betweenness centrality and Pagerank value of the nodes of the complex network and take the average value to obtain the comprehensive value of the node importance of the complex network that is less than the evaluation index of the controller status of the radar interval event investigation, that is, obtain the weight of the evaluation index of the controller status, which is used to reflect the global influence of the index in the entire complex network.

[0099] In one embodiment, the step S303 includes the following steps S3031-S3033:

[0100] S3031. The expression of the calculation model for standardizing the degree centrality of nodes in a complex network is:

[0101]

[0102] Among them, C′ D (i) represents the normalized value of the degree centrality of node i, C D (i) represents the degree centrality of node i, min{C D (j)} j∈[1，n] represents the minimum value of degree centrality in a complex network, max{C D (j)} j∈[1，n] Indicates the maximum value of degree centrality in a complex network;

[0103] S3032, using the calculation model to standardize the betweenness centrality and Pagerank value, respectively obtaining B′ i and PR′(i);

[0104] S3033. Calculate the average value of the standardized value to obtain the weight W of the evaluation index of the controller status. i , its calculation expression is:

[0105]

[0106] Understandably, by comprehensively considering degree centrality, betweenness centrality and Pagerank value, the importance of nodes in multiple dimensions in complex networks can be comprehensively evaluated. This multi-dimensional evaluation method can more accurately reflect the real influence of nodes than a single indicator. Standardizing centrality, betweenness centrality and Pagerank value ensures the numerical comparability of different indicators, eliminates the deviation caused by different dimensions or numerical ranges, and comprehensively considers the three dimensions to comprehensively evaluate the influence of controllers in the network. Degree centrality reflects the degree of direct connection between controllers and other controllers, betweenness centrality reveals the criticality of controllers as bridge nodes, and Pagerank value considers the impact of the overall structure and stability of the network on the importance of controllers. This multi-dimensional evaluation method can more accurately reflect the real influence of controllers than a single indicator. The level of the comprehensive value (weight) of node importance directly reflects the global influence of controllers in the entire complex network. Based on this evaluation result, managers can more reasonably allocate resources, such as training, incentives and job adjustments, to optimize the overall effectiveness of the controller team. This evaluation result also provides strong support for decision-making, helping managers make more informed and effective decisions when faced with complex situations.

[0107] S40, according to the weight of the controller status evaluation index, the danger index of the controller status is calculated by using the Z-Score normalization method;

[0108] In one embodiment, the step S40 includes the following steps S401-S403:

[0109] S401. Use the Z-Score normalization method to standardize all controller status evaluation indicators:

[0110]

[0111] Among them, μ i is the mean, σ i is the standard deviation, X′ is the standardized value of the indicator;

[0112] S402. Multiply the standardized indicator value of each indicator by the corresponding weight to obtain the weighted risk index value of each indicator:

[0113] R i =W i ×X′ i (14)

[0114] Among them, R i is the risk value of the ith indicator, W i is the weight of the i-th indicator, X′ i is the standardized value of the i-th indicator;

[0115] S403. Add the weighted risk index values ​​of all indicators to obtain the total risk index R T , its calculation expression is:

[0116]

[0117] Understandably, after the Z-Score normalization process, the status indicators of all controllers are converted to the same scale, so the status differences between different controllers can be compared more fairly and objectively. And because the Z-Score normalization method takes into account the distribution characteristics of the data (such as mean and standard deviation), it can more accurately reflect the actual degree of deviation of the controller's status indicators. This improvement in accuracy helps to more accurately assess the risk level of the controller, thereby providing managers with a more reliable basis for decision-making.

[0118] In one embodiment, the step S50 includes the following steps S501-S503:

[0119] S501. Customize the evaluation criteria based on the value of the danger index of the controller status evaluation indicator, as well as comprehensive historical data, industry standards and expert opinions;

[0120] S502, clarifying the specific definition and characteristics of each hazard level according to the hazard index, and determining preset thresholds for different hazard levels;

[0121] S503. Compare the hazard index with the preset threshold value, evaluate the current controller status, and provide a quantitative basis for risk management of events less than the specified radar interval, and provide relevant decision support, including timely issuing warnings, adjusting controller task allocation, or taking preventive measures.

[0122] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0123] In one embodiment, if Figure 3 As shown in FIG, the quantitative evaluation index system for the controller status in the less than radar interval event investigation includes 6 levels of dimensional indicators. Figure 4The complex network diagram shown in the figure shows that the nodes represent the causal factors that lead to the unsafe state of the controller, and the edges represent the logical relationship between the causal factors, which contains a total of 28 causal factors and 82 connecting edges. Through the topological characteristics analysis of this small complex network, it can be seen that: ① The average path length of the network is 2.08, indicating that it takes 2.08 steps for the causal factors to influence each other. ② The network diameter is 5, and the surface network requires at most 5 edges to connect any two factors in the network, which also shows that the network is a relatively closely connected network. ③ Busy airspace / high traffic volume D5, lack of on-site management E3, poor physical and mental state D2, hardware equipment failure / missing D6, untimely policy dissemination E5, and incomplete skills training E6. These six factors have a high out-degree, indicating that these factors are more likely to affect other nodes and are important factors that induce the unsafe state of controllers in the less than control interval event; Failure to effectively grasp the operating environment information C1, failure to effectively grasp the operating dynamic information C2, failure to effectively monitor and repeat C6, busy airspace with high traffic volume D5, and hardware equipment failure D6 have a high in-degree, indicating that these factors are affected by many other nodes in the network. They are the result driven by many other reasons in the causal network. ④ The top four with high betweenness centrality are: busy airspace / high traffic volume D5, hardware equipment failure / missing D6, lack of on-site management E3, and poor physical and mental state D2, indicating that these four factors play an important bridge role in causing the unsafe behavior of controllers. They are important nodes that control the flow of causal factors in the entire network and need to be focused on. ⑤ Improper action execution B3 and improper information preprocessing B1 have the highest Pagerank values, indicating that B3 and B1 are one of the potential risk sources in this complex network. The air traffic control department needs to strengthen the training and monitoring of controllers to ensure the standardization of their daily operations. ⑥ Busy airspace / high traffic volume D5, hardware equipment failure D6, lack of on-site management E3, poor physical and mental state D2, and improper action execution B3 are the top five factors that lead to the unsafe state of controllers in terms of node comprehensive value, involving objective environment, hardware equipment, management system, personnel status, operation execution and other aspects. The subsequent relevant control units can effectively reduce the frequency of unsafe behaviors of controllers and events less than the specified radar interval by improving and controlling these factors.

[0124] According to the above method, in a specific embodiment, on April 18, 2023, a Xiamen Airlines flight and a Kunming Airlines flight had an event of less than the specified interval in the Changzhou control area. The two aircraft triggered the TCASTA alarm. The controller of the take-off line tower of Changzhou Airport commanded the Xiamen Airlines aircraft to avoid, and then the two aircraft resumed to maintain the specified interval. After investigation, the incident was caused by the civil aviation controller of the take-off line tower not strictly complying with the regulations to use the flight progress sheet, mistaking the 3,900-meter altitude of the Xiamen Airlines flight for 3,300 meters, and mistakenly issuing a descending altitude instruction; the controller left for the take-off line tower on the other side due to the need for preparatory work for changing the runway, and transferred the responsibilities of the seat to the notification and coordination seat; the controller of the notification and coordination seat failed to perform the transferred monitoring seat responsibilities due to the heavy workload of coordination, and failed to discover the erroneous instruction in time, which eventually led to the conflict between the flight and the Kunming Airlines flight flying at an altitude of 3,600 meters on the same route, which was less than the specified interval. The investigation also found that the heavy workload of the controllers due to the airport's non-stop construction, the co-operation of military and civil aviation, the replacement of runways, and the complex control operation environment were also contributing factors to the occurrence of the incident. The minimum interval between the two aircraft in this incident was: 60 meters vertically, 19 kilometers horizontally, less than 1 / 3 but not less than 1 / 5 of the prescribed interval in the vertical direction, and less than 1 / 5 of the prescribed interval in the horizontal direction. According to the "Civil Aircraft Sign Registration Division Method" (AC-395-AS-01), this incident constitutes a general transport aviation sign of the airport civil aviation control responsibility. However, the controller's status in this incident was not quantitatively evaluated. Therefore, this section analyzes the investigation report, explores the human factors in the event of less than the control interval, and quantitatively scores them according to the actual situation and the risk index (as shown in Table 2 below). Finally, the controller's status risk index score in this incident is 8.053 points, which provides ideas for the quantitative analysis of the controller's status.

[0125] Table 2 Calculation of controller status hazard index in the event of less than controlled interval

[0126]

[0127] In one embodiment, a quantitative assessment system for controller status in less than radar interval event investigation is provided, and the quantitative assessment system for controller status in less than radar interval event investigation corresponds to the quantitative assessment method for controller status in less than radar interval event investigation in the above embodiment. Figure 2 As shown, the system comprises:

[0128] An extraction module, used to extract causal keywords based on the investigation report of the less than radar interval incident;

[0129] Improved HFACS model construction module, used to construct an improved HFACS model based on the HFACS model and causal keywords, and to construct a controller status evaluation index system based on the improved HFACS model;

[0130] A weight determination module, used to determine the weight of the controller status evaluation index based on the controller status evaluation index system;

[0131] A danger index calculation module is used to calculate the danger index of the controller status evaluation index using a Z-Score normalization method according to the weight of the controller status evaluation index;

[0132] The evaluation module is used to formulate the evaluation criteria of the controller status based on the danger index of the controller status evaluation index.

[0133] The specific limitations of the quantitative assessment system for the state of air traffic controllers for investigation of events with less than radar intervals can be found in the limitations of the quantitative assessment method for the state of air traffic controllers for investigation of events with less than radar intervals mentioned above, which will not be repeated here. Each module in the above-mentioned quantitative assessment system for the state of air traffic controllers for investigation of events with less than radar intervals can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0134] In one embodiment, a controller is provided, comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein when the processor executes the computer-readable instructions, a method for quantitatively evaluating the state of a controller according to the investigation of the less-than-radar interval event is implemented. Further, the controller comprises a processor, a memory, a network interface, and a database connected via a system bus. The processor of the controller is used to provide computing and control capabilities. The memory comprises a readable storage medium and an internal memory. The readable storage medium stores an operating system, computer-readable instructions, and a database. The internal memory provides an environment for the operation of the operating system and computer-readable instructions in the readable storage medium. When the computer-readable instructions are executed by the processor, a method for adjusting the height of a suspension is implemented. The readable storage medium provided in this embodiment comprises a non-volatile readable storage medium and a volatile readable storage medium.

[0135] In one embodiment, a computer-readable storage medium is provided, the computer-readable storage medium storing computer-readable instructions, the computer-readable instructions being executed by a processor to implement the controller status quantitative assessment method for the less than radar interval event investigation in the above embodiment, to avoid repetition, it is not described here. The computer-readable storage medium may be non-volatile or volatile.

[0136] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through computer-readable instructions, and the computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When the computer-readable instructions are executed, they may include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0137] The above are merely embodiments of the method and system for quantitatively evaluating the status of controllers for investigating events less than radar intervals of the present invention, and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for quantitatively evaluating the state of a controller for investigation of less than radar interval events, characterized in that: include: S10, extracting causal keywords based on the investigation report of the less than radar interval event; S20, constructing an improved HFACS model based on the HFACS model and the causal keywords, and constructing a controller status evaluation index system based on the improved HFACS model; the step S20 includes: S201. Replace the specific indicators of each level of the HFACS model with the causal keywords, and replace the unsafe supervision indicators in the HFACS model, such as insufficient supervision, inappropriate operation plan, failure to correct problems, and supervision violations, with unreasonable team matching, fatigue / overtime work, lack of on-site management, regular spot checks that are superficial, untimely policy dissemination, and incomplete skills training; S202, adding an unsafe behavior performance indicator layer to the unsafe behavior indicator layer of the HFACS model; The unsafe behavior indicator layer includes improper information processing, improper decision-making and improper action execution; the unsafe behavior performance indicator layer includes failure to effectively grasp environmental information, failure to effectively grasp operation dynamics, unreasonable control plan, failure to adjust command in response to operation conflicts, issuance of wrong control instructions, failure to effectively monitor repetition and failure to ensure consistency between movement status and process list, so as to form an improved HFACS model; S203, sorting out the correlation relationship between the controller status evaluation indicators of each level of the improved HFACS model that replaces the causal keywords and adds the safety behavior performance indicator layer, obtaining the causal chain of the less than radar interval event, and constructing a controller status evaluation indicator system; S30, determining the weight of the controller status evaluation index based on the controller status evaluation index system; S40, according to the weight of the controller status evaluation index, the danger index of the controller status evaluation index is calculated by using the Z-Score normalization method; S50. Based on the hazard index of the controller status evaluation index, formulate the evaluation criteria for the controller status.

2. The method for quantitatively evaluating the controller status for less than radar interval event investigation according to claim 1, characterized in that: The step S10 comprises: S101, preprocessing the less than radar interval event investigation report, wherein the preprocessing includes cleaning the style of the investigation report and removing stop words, punctuation marks and irrelevant characters; S102, segmenting the text into words or phrases, removing common and meaningless words, and creating or obtaining a stop word list; S103, after filtering the stop words in the stop word list in the text, use the non-stop words adjacent to the text as candidate keywords, calculate the keyword score based on the word frequency and co-occurrence frequency of the keywords, sort all the candidate keywords according to the score, and select the keyword with the highest score as the final causal keyword of the controller status assessment indicator of the less than radar interval event.

3. The method for quantitatively evaluating the controller status for less than radar interval event investigation according to claim 1, characterized in that: The step S30 comprises the following steps: S301, constructing a complex network based on the causal chain in the controller status evaluation index system, and performing topological feature analysis on the complex network; the complex network includes nodes and edges, the nodes represent the controller status evaluation indicators of each level of the improved HFACS model, and the edges represent the connection relationship between the nodes; S302, performing importance analysis on the nodes of the complex network according to the result of topological feature analysis of the complex network, wherein the importance analysis includes analysis of degree centrality for determining the degree of connection between a node and other nodes, betweenness centrality for determining whether it is a key bridge node in the complex network, and Pagerank value for determining the stability of the network; S303. Standardize the degree centrality, betweenness centrality and Pagerank value of the nodes of the complex network and take the average value to obtain the comprehensive value of the node importance of the complex network that is less than the evaluation index of the controller status of the radar interval event investigation, that is, obtain the weight of the evaluation index of the controller status, which is used to reflect the global influence of the index in the entire complex network.

4. The method for quantitatively evaluating the controller status for less than radar separation event investigation according to claim 1, characterized in that: The step S50 comprises the following steps: S501. Customize the evaluation criteria based on the value of the danger index of the controller status evaluation indicator, as well as comprehensive historical data, industry standards and expert opinions; S502, clarifying the specific definition and characteristics of each hazard level according to the hazard index, and determining preset thresholds for different hazard levels; S503. Compare the hazard index with the preset threshold value, evaluate the current controller status, and provide a quantitative basis for risk management of events less than the specified radar interval, and provide relevant decision support, including timely issuing warnings, adjusting controller task allocation, or taking preventive measures.

5. A quantitative assessment system for controller status for less than radar interval event investigation that implements the quantitative assessment method for controller status according to claim 1, characterized in that: include: An extraction module, used to extract causal keywords based on the investigation report of the less than radar interval incident; Improved HFACS model construction module, used to construct an improved HFACS model based on the HFACS model and causal keywords, and to construct a controller status evaluation index system based on the improved HFACS model; A weight determination module, used to determine the weight of the controller status evaluation index based on the controller status evaluation index system; A danger index calculation module is used to calculate the danger index of the controller status evaluation index using a Z-Score normalization method according to the weight of the controller status evaluation index; The evaluation module is used to formulate the evaluation criteria of the controller status based on the danger index of the controller status evaluation index.

Images