Pilot fatigue monitoring method

Abstract

The invention discloses a pilot fatigue monitoring method, which comprises the following steps: generating an electronic checklist, in which checklist items for fatigue monitoring are arranged randomly and are located at the last of the electronic checklist; the pilot executes the electronic checklist item, and for the checklist item used for fatigue monitoring, an eye tracker is used for recording checking process data of the pilot; and after the pilot executes the electronic inspection list task, counting inspection process data of the fatigue monitoring items recorded by the eye tracker, comparing the inspection process data with reference data of the pilot, and judging the fatigue monitoring result. The fatigue of the pilot can be found, so that the flight safety is improved.

Description

Technical Field

[0001] The present invention relates to an evaluation method in the field of aviation, specifically a method for monitoring pilot fatigue, which is used to monitor and evaluate the fatigue state of pilots. Background Art

[0002] Although currently large airliners adopt a two-pilot crew, the concept of single-pilot operation has also been proposed for many years. When operating alone, since there is only one pilot, the mental state of this pilot is closely related to flight safety. Therefore, it is necessary to monitor and evaluate the fatigue state of the pilot. For this purpose, in the present invention, when a pilot performs the pre-flight checklist tasks during single-pilot operation, the execution situations of some checklist items are extracted as the basis for fatigue monitoring, and the current fatigue monitoring and evaluation result of the pilot is given by comparing the data of the current checklist execution situation of the pilot with the historical execution situation. If fatigue is detected, relevant measures can be initiated to relieve the pilot's fatigue and thus improve flight safety. Summary of the Invention

[0003] The object of the present invention is to provide a method for monitoring pilot fatigue. In the single-pilot operation mode, when the pilot performs the pre-flight checklist tasks, some checklist items are extracted as the basis for fatigue monitoring. An eye tracker is used to monitor the working state of the pilot during these checklist tasks, and the current fatigue monitoring and evaluation result of the pilot is given by comparing the data of the current checklist execution situation of the pilot with the historical execution situation. If fatigue is detected, relevant measures can be initiated to relieve the pilot's fatigue and thus improve flight safety.

[0004] The object of the present invention is achieved through the following technical solutions:

[0005] A method for monitoring pilot fatigue includes the following steps:

[0006] Step 1: Generate an electronic checklist, where the arrangement order of the checklist items for fatigue monitoring is randomly arranged and is located at the end of the electronic checklist;

[0007] Step 2: The pilot executes the electronic checklist items, and for the checklist items for fatigue monitoring, an eye tracker is used to record the inspection process data of the pilot;

[0008] Step 3: After the pilot finishes executing the electronic checklist tasks, the inspection process data of the fatigue monitoring items recorded by the eye tracker is statistically analyzed, and compared with the baseline data of the pilot to evaluate the fatigue monitoring result of this time.

[0009] Preferably, in Step 1, the arrangement order of the checklist items for fatigue monitoring is randomly arranged based on the distance weight of the inspection target, and the steps are as follows:

[0010] (1). Let the index values of the first and last items of the checklist for fatigue monitoring be A and B, (A < B);

[0011] (2). Starting from the last item of the checklist for fatigue monitoring, generate a random number J ranging from A to the index value of the current item;

[0012] (3). Let the current position index be X. If the distance between the inspection target corresponding to the item at index X + 1 and the inspection target corresponding to the item at index J is greater than 30 cm, then exchange the contents of the two items at indices X and J. Otherwise, continue to generate the random number J, and then continue to compare the distances. If the number of repetitions is greater than 10(B - A) times and still not successful, then use the random number value of the last J to exchange the contents of the two items;

[0013] (4). Repeat steps (2) to (3), moving forward step by step until reaching the item at index A.

[0014] Preferably, in step two, when the pilot executes the checklist inspection item, when reaching the fatigue monitoring item, the fatigue monitoring item is displayed at the bottom of the page. After completing this fatigue monitoring item, the screen scrolls up one item, and the next fatigue monitoring item appears at the bottom of the page, preventing the pilot from seeing the next item in advance when executing the current item.

[0015] Preferably, in step two, the inspection process data includes:

[0016] Using an eye tracker to monitor and record the moment when the pilot leaves the fatigue monitoring item after reading it, denoted as T Bi 、

[0017] The moment when the pilot finds the target corresponding to this fatigue monitoring item, denoted as T Fi ,

[0018] The saccade path of the pilot searching for this target, denoted as L Ri 。

[0019] Preferably, the pilot's reference data refers to the inspection process data of the fatigue monitoring items recorded by an eye tracker when the pilot performs the same checklist task during daily training using a D - level simulator under the condition of sufficient sleep and good condition; at least 10 times of data collection are carried out, and data statistics are performed. When performing statistics, the execution conditions of the checklist items with the same content are statistically analyzed, including the statistical average time for target search, denoted as D i , D i =(T Fi -T Bi ) / M, where M is the number of data collections, and the statistical average path for target search, denoted asL i , L i = L Ri / M.

[0020] Preferably, in step three, when comparing the pilot's reference data, a scoring system is used to give the fatigue monitoring result. Let the initial fatigue score be S, S = 0, the initial number of fatigue monitoring items be M. For the i-th fatigue monitoring item,

[0021] If L Ri / L i > 1.5, and T Ri / D i < 1, then the inspection of this item may be abnormal, the result is not recorded, and the value of M is reduced by 1;

[0022] If L Ri / L i < 1.5, and T Ri / D i < 1.5, then S is increased by 1;

[0023] For all fatigue monitoring items, the above statistics are carried out. If the final statistical result S / M < 0.5, it is considered that the pilot is in a fatigue state.

[0024] The beneficial effects of the present invention are as follows:

[0025] 1. The steps of the present invention are simple, the design is reasonable, and the implementation and use operations are convenient.

[0026] 2. The present invention generates an electronic checklist. The content of the items in the electronic checklist is the same as that in the paper checklist, but the arrangement order of the items for fatigue monitoring is randomly arranged, and the items for fatigue monitoring are located at the end of the electronic checklist;

[0027] 3. The present invention enables the pilot to execute the inspection items of the electronic checklist. For the fatigue monitoring items, an eye tracker is used to record the inspection process data of the pilot, while for the non-fatigue monitoring items, no record is made; after the pilot finishes executing the electronic checklist task, the inspection process data of the fatigue monitoring items recorded by the eye tracker is statistically analyzed, and the fatigue monitoring result is given after comparing with the pilot's reference data.

[0028] In summary, the technical solution of the present invention is reasonably designed. By using the generated electronic checklist, fatigue monitoring is carried out when the pilot performs the task of the electronic checklist, avoiding the additional steps of fatigue monitoring and saving the pre-flight preparation time. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1Schematic diagram of the saccade path for a pilot fatigue monitoring method of the present invention. Detailed implementation manners

[0030] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0031] A pilot fatigue monitoring method includes the following steps:

[0032] Step 1: Generate an electronic checklist. The entry content of the electronic checklist is the same as that of the paper checklist, but the arrangement order of the checklist entries for fatigue monitoring adopts a random arrangement based on the distance weight of inspection targets, and the checklist entries for fatigue monitoring are located at the end of the electronic checklist.

[0033] In Step 1, the electronic checklist is a checklist displayed using an electronic display, that is, the paper checklist is digitized and displayed on the screen. The checklist entries for fatigue monitoring are mainly the checklist entries that require the pilot to observe and confirm the status, and only one target status needs to be observed. The content of these entries can be selected and customized from the paper checklist as needed. These checklist entries that can be used for fatigue monitoring can be, but are not limited to, such as "The electric hydraulic pump switch is in the OFF position", "The battery switch is in the ON position and the cover is properly closed", etc. For example, for "The landing gear handle is in the down position and three green lights are on", it is necessary to observe the landing gear handle and three green lights, so it should not be directly used as a checklist entry for fatigue monitoring. However, after this checklist is decomposed into two checklist entries, it can be used as a checklist entry for fatigue monitoring, and after decomposition, it is "The landing gear handle is in the down position" and "Three green lights of the landing gear indicator are on".

[0034] In Step 1, the arrangement order of the checklist entries for fatigue monitoring adopts a random arrangement based on the distance weight of inspection targets, which means that it is necessary to ensure that the distance between the inspection targets required for two adjacent checklist entries for fatigue monitoring is not less than 30 cm as much as possible. The improved Fisher-Yate shuffling algorithm can be used, but is not limited to it. The improved Fisher-Yate shuffling algorithm is as follows:

[0035] The specific steps of the Fisher-Yates shuffling algorithm are as follows:

[0036] (1). Let the index values of the first and last checklist entries for fatigue monitoring be A and B, (A < B);

[0037] (2). Starting from the last entry of the checklist entries for fatigue monitoring, generate a random number ranging from A to the current entry index value, denoted as J;

[0038] (3). Let the current position index be X. If the distance between the inspection target corresponding to the entry at index X + 1 (X + 1 ≤ B, and if X equals B, there is no need to compare distances) and the inspection target corresponding to the entry with index value J is greater than 30 cm, then exchange the contents of the two entries with indices X and J. Otherwise, continue to generate a random number J, and then continue to compare distances. If the number of repetitions is greater than 10(B - A) times and still not successful, then use the random number value of J in the last time to exchange the contents of the two entries;

[0039] (4). Repeat steps (2) to (3) step by step forward until reaching the entry with index A.

[0040] Suppose the checklist entries for current fatigue monitoring have been placed at the end of the electronic checklist. Assume the index A of the first fatigue monitoring entry is 50, and the last index B is 54. At this time, there are 5 checklist entries for fatigue monitoring. First, randomly arrange the index 54. At this time, X = 54, and generate a random number between 50 and 54. Suppose the random number is J = 51. Since it is the last index currently, that is, X = 54, there is no need to compare distances, and directly exchange the checklist contents corresponding to indices X = 54 and J = 51; Next, process the index X = 53, generate a random number between 50 and 53, and set the random number J = 50. If the distance between the target to be observed corresponding to the entry with index J = 50 and the target to be observed corresponding to the checklist entry with index X + 1 = 54 is greater than 30 cm, exchange the checklist contents corresponding to the current indices X = 53 and J = 50. If it is less than 30 cm, then it is necessary to generate a random number between 50 and 53 again, and then check the distance. If the exchange condition is still not met after repeating 10×(54 - 50) = 40 times, then select the value of J in the last time for exchange. Repeat this process until the index X = 50 is processed.

[0041] Step 2: The pilot executes the electronic checklist entries. When the fatigue monitoring entry appears, it should be displayed at the bottom of the page, and measures should be taken to prevent the pilot from seeing the content of the next checklist item in advance during the execution. For the fatigue monitoring entry, an eye tracker is required to record the pilot's inspection process data, and for non-fatigue monitoring entries, no recording is done.

[0042] In Step 2, the pilot executes the electronic checklist inspection entries. When the pilot executes to the fatigue monitoring entry during the execution process, the fatigue monitoring entry needs to be displayed at the bottom of the page. After each execution of this fatigue monitoring entry, the screen scrolls up one entry, and the next fatigue monitoring entry appears at the bottom of the page, preventing the pilot from seeing the next item in advance when executing the current item;

[0043] In step 2, the eye tracker is an instrument that can be used to track and record human eye activities. General commercial eye trackers have the function of obtaining the gaze target and scanning path of the human eye.

[0044] In step 2, the eye tracker is used to record the pilot's inspection process data for the fatigue monitoring items. This means that the eye tracker is used to monitor and record the time when the pilot looks at and reads the fatigue monitoring item and then leaves the item, which is recorded as T Bi The time when the pilot finds the target corresponding to the fatigue monitoring item is recorded as T Fi , and the pilot's glance path from the checklist item to the target, denoted as L Ri , where i = 1, 2, 3, ... N, is the corresponding i-th fatigue monitoring item, and N is the number of fatigue monitoring items;

[0045] Step 3: After the pilot completes the electronic checklist task, the inspection process data of the fatigue monitoring items recorded by the eye tracker are collected and compared with the pilot's baseline data, and the fatigue monitoring results are evaluated using a scoring system.

[0046] In step 3, the pilot's baseline data refers to the fatigue monitoring item check process data recorded by the eye tracker when the pilot performs the same checklist task during daily training using a Level D simulator with adequate sleep and good condition; at least 10 data collections are performed and data statistics are performed. When performing statistics, it is necessary to count the execution of the checklist items with the same content, mainly the average target search time statistics. For item i, the average search time is recorded as D i , D i =(T Fi -T Bi ) / M, where M is the number of data collection times and the average path statistics of the search target, recorded as L i , L i =L Ri / M, calculate L i When the pilot hesitates, the path that is too long needs to be manually eliminated. The elimination method is to sort the path lengths obtained multiple times and take the median path length as the benchmark. If a certain scanning path is longer than twice the benchmark path, it will be eliminated. At the same time, the search time corresponding to the record should also be eliminated.

[0047] In step 3, the statistics of the inspection process data of the fatigue monitoring items recorded by the eye tracker are calculated, which means that for each fatigue monitoring item, the corresponding search time is calculated, which is recorded as T Ri , T Ri =T Fi-T Bi , while calculating the search path L Ri . The search path is the saccade path where the pilot's eyes leave the checklist item and start searching for the target, and it is the path on the image obtained by mapping the saccade path recorded by the eye tracker onto a unified cockpit picture. As shown in the appendix Figure 1 . In the figure, the checklist is a schematic diagram of an electronic checklist, and the red path in the figure is an example of the path between the pilot's line of sight leaving the fatigue monitoring item of the checklist and the required observation target after being mapped onto the cockpit image

[0048] In step three, when comparing the pilot's reference data and rating the fatigue monitoring result using a scoring system, it means setting the initial fatigue score as S, S = 0, the initial number of fatigue monitoring items as M. For the i-th fatigue monitoring item

[0049] If L Ri / L i > 1.5, and T Ri / D i < 1, (L Ri is the saccade path length when this item is executed L i is the reference path length of this item, T Ri is the search time when this item is executed D i is the reference search time), then the inspection of this item may be abnormal, the result is not recorded, and the value of M is decreased by 1;

[0050] If L Ri / L i < 1.5, and T Ri / D i < 1.5, then S is incremented by 1;

[0051] For all fatigue monitoring items, the above statistics are performed. If the final statistical result S / M < 0.5, it is considered that the pilot is in a fatigued state

[0052] It can be understood that for those of ordinary skill in the art, equivalent substitutions or changes can be made according to the technical solution of the present invention and its inventive concept, and all such changes or substitutions should fall within the protection scope of the claims appended to the present invention

Claims

1. A pilot fatigue monitoring method, characterized in that The following steps are involved: Step 1: Generate an electronic checklist, where the checklist items for fatigue monitoring are arranged in a random order and are placed at the end of the electronic checklist; Step 2: The pilot performs the electronic checklist items. For the checklist items used for fatigue monitoring, the pilot's checklist process data is recorded using an eye tracker. Step 3: After the pilot completes the electronic checklist task, the inspection process data of the fatigue monitoring items recorded by the eye tracker are collected and compared with the pilot's baseline data to evaluate the fatigue monitoring results.

2. A pilot fatigue monitoring method according to claim 1, characterized in that In step 1, the order of fatigue monitoring checklist items is randomly arranged based on the weight of the inspection target distance. The steps are as follows: (1) Assume that the index values of the first and last items in the fatigue monitoring checklist are A and B, (A <B); (2) Starting from the last item in the fatigue monitoring checklist, generate a random number ranging from A to the index value of the current item, denoted as J; (3) Let the current position index be X. If the distance between the inspection target corresponding to the entry with index X+1 and the inspection target corresponding to the entry with index J is greater than 30 cm, then swap the contents of the entry with index X and the entry with index J. Otherwise, continue to generate a random number J and then continue to compare the distances. If the number of repetitions is greater than 10 (BA) and still unsuccessful, then use the last random value of J to swap the contents of the two entries. (4) Repeat steps (2) to (3), proceeding step by step until the entry with index A is reached.

3. A pilot fatigue monitoring method according to claim 1, characterized in that In step 2, when the pilot executes the electronic checklist checklist items, when executing the fatigue monitoring item, the fatigue monitoring item is displayed at the bottom of the page. After executing the fatigue monitoring item, the screen scrolls up one item and the next fatigue monitoring item appears at the bottom of the page, preventing the pilot from seeing the next item in advance while executing the current item.

4. A pilot fatigue monitoring method according to claim 1, characterized in that In step 2, the inspection process data includes: Use an eye tracker to monitor and record the time when the pilot leaves the fatigue monitoring item after reading it, and record it as T Bi 、 The time when the pilot finds the target corresponding to the fatigue monitoring entry is recorded as T Fi , The pilot's scanning path to search for the target is recorded as L Ri .

5. A pilot fatigue monitoring method according to claim 4, characterized in that In step 3, the pilot's baseline data refers to the fatigue monitoring item check process data recorded by the eye tracker when the pilot performs the same checklist task during daily training using a Level D simulator with adequate sleep and good condition; at least 10 data collections are performed and data statistics are performed. When statistics are performed, the execution status of the checklist items with the same content is counted, including the average target search time statistics, which is recorded as D i , D i =(T Fi -T Bi ) / M, where M is the number of data collection times and the average path statistics of the search target, recorded as L i , L i =L Ri / M.

6. A pilot fatigue monitoring method according to claim 5, characterized in that In step 3, when comparing the pilot's baseline data, a scoring system is used to give the fatigue monitoring results. Let the initial fatigue score be S, S = 0, the number of initial fatigue monitoring items be M, and for the i-th fatigue monitoring item, If L Ri / L i >1.5, and T Ri / D i <1, the check of this item may be abnormal, the result is not recorded, and the M value is reduced by 1; If L Ri / L i <1.5, and T Ri / D i <1.5, then S increases by 1; The above statistics are performed for all fatigue monitoring items. If the final statistical result S / M is less than 0.5, the pilot is considered to be in a fatigue state.

Images