JP5454395B2 - Passenger conveyor user fall detection device - Google Patents

Description

  The present invention relates to a user fall detection device for passenger conveyors such as escalators and electric roads.

  In a passenger conveyor, when a user (passenger) gets off from a passenger conveyor in operation to a boarding gate, there is a concern that the passenger conveyor may fall down accidentally. Even if a user falls down accidentally at the entrance / exit, it is usually impossible to detect the fall of the user, so that improvement is desired.

  In the conventional passenger conveyor, a laser scan sensor is installed in the vicinity of the passenger conveyor entrance and exit, and the movement of the user on the plane coordinates is measured by the laser scan sensor, and the user's measurement measured by the laser scan sensor is performed. When the moving speed (residence status information at the boarding / exiting section) falls below a predetermined value, the voice synthesizer broadcasts an alert, and when the stay or congestion of the customer is prolonged, the frequency and voltage generated by the inverter device are controlled. Thus, a passenger conveyor safety device has been proposed in which the speed of the drive motor is reduced or stopped (see, for example, Patent Document 1). This passenger conveyor safety device creates a coordinate map of continuous points on the surface of an object to be detected (passenger) by a laser scan sensor installed in the vicinity of a passenger conveyor entrance and exit, and there are no users who have been measured in advance. Overlay with the coordinate data map in the state and remove the coordinate data of the fixed obstacle, thereby obtaining the contour data of the passenger by the current plane coordinate data map on the passenger conveyor. And from the contour data of the passenger, the coordinates of the center point of the passenger are calculated, the moving speed of each detection object (passenger) is calculated, and when the moving speed of the detection object falls below a predetermined threshold, Even when the stoppage or stay of the passenger is detected, and even when the stoppage or stay of the passenger is not detected, if the detection object multiple detection threshold is exceeded, it is determined that the passenger conveyor is getting on and off.

JP 2008-303057 A

  In a conventional passenger conveyor safety device, a laser scan sensor captures the center point from the detected contour data (shape) of the object to be detected (user), and stays or is congested based on the number of center points and the moving speed. When the user (passenger) falls down, for example, even if the head is near the entrance / exit, the foot may be outside the scan area. In such a state, the center point is accurately obtained. In some cases, the fall cannot be detected. Even if the center point is obtained, there is a case where it is determined that there is no problem and the alarm is not sounded, except for the entrance / exit. In addition, when a fallen user moves his hand, such as flapping or making a slight gesture, the center point is captured by the hand part, and it moves moderately. There is a possibility that the fall cannot be detected. Further, if the laser reaches far away, it may malfunction due to stay away from the passenger conveyor.

  The present invention has been made to solve the above-described problems, and even when a fallen person flutters, the fallen state can be reliably detected without erroneously recognizing the congestion state during staying. A user fall detection device for a passenger conveyor is provided.

  In the user fall detection device for a passenger conveyor according to the present invention, a passenger conveyor and a boarding / alighting board that is installed at each boarding / exiting gate that is a landing area and a landing area of the passenger conveyor, and a comb board is provided on the step side of the passenger conveyor. Floor board, approach passage provided on the side of the anti-combination board for boarding / exiting board, and approaching passage for passengers to approach, and boarding / exiting floor board, A scanning distance sensor that emits a laser beam in a horizontal plane in the scanning range including the approach passage, and the scanning distance sensor scans the surface of the fallen person and integrates the area of the area hitting the body part with time. When the value exceeds a certain value, it is determined that the device is in a fallen state.

  In addition, the scan type distance sensor has a threshold value of about 150 (square centimeter x second) obtained by integrating the area of the area corresponding to the body part over time, so that the situation at the time of dwelling and the situation at the time of falling over can be distinguished. Is.

  According to the present invention, even when a fallen person flutters, the fallen state can be reliably detected without erroneously recognizing the crowded state during staying.

It is a side view which shows the whole schematic structure provided with the user fall detection apparatus of the passenger conveyor in the basic example of this invention. It is a top view which shows the user fall detection apparatus of the passenger conveyor in the basic example of this invention. It is a perspective view of the entrance / exit part which shows the user fall detection device of the passenger conveyor in the basic example of this invention. It is explanatory drawing which shows the scanning range of the scanning type distance sensor used for the user fall detection apparatus of the passenger conveyor in the basic example of this invention. It is explanatory drawing which shows the calculation image for the scanning range of the scanning type distance sensor used for the user fall detection apparatus of the passenger conveyor in the basic example of this invention, and fall. It is explanatory drawing which shows a mode that a fall state and a staying state are analyzed by the user fall detection apparatus of the passenger conveyor in the basic example of this invention. It is a flowchart for demonstrating the fall detection algorithm of the user fall detection apparatus of the passenger conveyor in the basic example of this invention. It is explanatory drawing which shows the condition at the time of the residence of a passenger conveyor in simulation. It is a characteristic view which shows the condition at the time of a residence of a passenger conveyor by a shadow area analysis result. It is a characteristic view which shows the result of time-integrating the shadow area of the situation at the time of the residence of a passenger conveyor. It is explanatory drawing which shows the condition at the time of the overturn flutter of a passenger conveyor simulated. It is a characteristic view which shows the condition at the time of the flipping over of a passenger conveyor by a shadow area analysis result. It is a characteristic view which shows the result of time-integrating the shadow area of the situation at the time of the flipping over of the passenger conveyor. It is explanatory drawing which shows a mode that the fall area is analyzed by time-integrating a shadow area | region by the user fall detection apparatus of the passenger conveyor in Example 1 of this invention. It is a flowchart for demonstrating a fall detection algorithm from the shadow area | region of the user fall detection apparatus of the passenger conveyor in Example 1 of this invention.

Basic example

  FIG. 1 is a side view showing an overall schematic configuration of a passenger conveyor user fall detection device according to a basic example of the present invention, FIG. 2 is a plan view showing the passenger fall user fall detection device, and FIG. 3 is a passenger conveyor. FIG. 4 is an explanatory view showing a scanning range of a scanning distance sensor used for a user fall detection device of a passenger conveyor and a calculation image for detecting the fall, FIG. 6 is an explanatory diagram showing a state in which a fall state and a staying state are analyzed by a user fall detection device for a passenger conveyor, and FIG. 7 is a flowchart for explaining a fall detection algorithm of the user fall detection device for a passenger conveyor.

  1 to 3, reference numeral 1 denotes a passenger conveyor, which is assumed to be down. 2 is the entrance / exit of the first floor where the passenger conveyor 1 gets off, 3 is the entrance / exit of the second floor where the passenger conveyor 1 is placed, and 4 is the floor board for getting on / off the passenger conveyor installed at each of the entrances 2, 3 The passenger enters the passenger conveyor 1 step from the boarding board 4 or gets off the passenger conveyor 1 step. 5 is a comb plate provided at the front end portion of the step board of the boarding board 4 and 6 is provided on the side of the anti-comb board 5 of the boarding board 4 for passengers to approach the boarding board 4 of the passenger conveyor 1. This is an approach passage. Here, the case where the user (passenger) falls on the boarding board 4 for boarding / alighting of the entrance / exit 2 of the 1st floor which is a landing is shown. 7 is a laser scan sensor installed near one side of the first floor entrance / exit 2 and near one side of the second floor entrance 3 so as not to obstruct passenger traffic. As shown in FIG. 3, the scanning distance sensor may be installed in a column. Further, it may be a set with a laser scan camera, capture a fall detection video with a small camera, save the fall detection video, or send it to an administrator. As shown in FIG. 2, the scan type distance sensor 7 emits a laser beam in the horizontal direction from the vicinity of one side of the entrances 2 and 3 and rotates the optical axis of the laser in the vertical direction to center the sensor. The distance in the horizontal direction is measured. The scanning range 7a of the laser beam of the scanning distance sensor 7 installed near one side of the entrance / exit 2 is that passengers approach the entrance / exit floor plate 4 at the entrance / exit 2 on the first floor. Are scanned so as to include the range of the approach passage 6, the floor board 4 for getting on and off, and the comb board 5. Further, the scanning range 7a of the laser beam of the scanning distance sensor 7 installed near one side of the entrance / exit 3 is a comb board 5, an entrance / exit floor board 4 and passengers at the entrance / exit 3 on the second floor. Scanning is performed so as to include the range of the approach passage 6 for approaching the boarding board 4. That is, the scanning range 7a of the laser beam includes the approach passage 6 for approaching the boarding board 4 from a distant position as well as the boarding board 4 and the comb board 5 provided in the vicinity of the boarding doors 2 and 3. It is scanned so as to include. Reference numeral 8 denotes a processing device connected to the scan type distance sensor 7. The scanning distance sensor 7 and the processing device 8 are installed, and the size or length of the object is measured by the scanning distance sensor 7. The distance for each angle measured by the scanning distance sensor 7 is accumulated, and the standard deviation with respect to the temporal distance change and the standard deviation of the standard deviation are calculated for the data in the detection area. When the standard deviation is equal to or smaller than the threshold value and the standard deviation of the standard deviation is equal to or smaller than the threshold value, the surface size (length) of the object in the region is calculated. And when the state whose surface size (length) is more than a threshold lasts more than predetermined time, it judges with a user (passenger) falling state. As shown in FIG. 4, the scan type distance sensor 7 measures the distance R with respect to the object at an angular pitch of 0.36 ° (360 ° is divided into 1024) within a range of ± 120 °. The unit of measurement is millimeters (mm). The scanning period is 100 ms. For the sake of simplicity, FIG. 5 is illustrated at a pitch of 15 °, and the ◯ marks are measurement points. The dotted frame in FIG. 5 indicates the detection area, and the points A, B, and H are not used for calculation because they are outside the detection area. The black arrow in FIG. 5 indicates the standard deviation of the distance fluctuation of the measurement point in the past 1-2 seconds, and the white arrow indicates the range of the standard deviation of the past 2-8 points of the standard deviation. The standard deviation of the standard deviation is obtained by further taking the standard deviation in the time axis direction in order to see the temporal variation width of the standard deviation obtained previously. Since the points C, D, E, F, and G in FIG. 5 are within the detection area, they are used for the calculation. The arcs further protruding on both sides between the points D-E in FIG. 5, the arcs protruding on both sides of the point F, and the arcs protruding on both sides of the point G are object surface sizes. Here, the boundary processing of the detection area will be described. The point G may be out of the detection area in the past 1-2 seconds. In this case, the points outside the detection area are not added to the calculation. When both the standard deviation, the standard deviation of the standard deviation, or the selected side is below the threshold, the object surface size is calculated. If the points D, E, F, and G are equal to or less than the threshold value, the lengths of the respective arcs are calculated and the sum is set as the object surface size. In FIG. 6, the horizontal axis is the standard deviation value (or the standard deviation value of the standard deviation), and the vertical axis is the time when the predetermined object surface size is exceeded. As is clear from FIG. 6, since the fall state has a long duration and the stay state has a short duration, if there is an interval between the fall state and the stay state, it can be correctly determined using this interval as a margin. To do. Further, a place having the widest interval may be set as a determination threshold value. According to the analysis result based on the actual standard deviation, it was found that the fall could be correctly determined when the standard deviation threshold was 60 to 80 mm, the object surface size threshold was 400 to 600 mm, and the duration was 5 to 10 seconds. However, in this case, if the fluttering occurs in front of the scan-type distance sensor 7, it is not determined that the body has fallen. (Battery collapse: move limbs at 30 cm reciprocation / second). Moreover, according to the analysis result by the standard deviation of the actual standard deviation, the standard deviation threshold of the standard deviation is 20 to 40 mm, the object surface size threshold is 300 to 500 mm, and the duration is 1-2 seconds. It can be determined, and in this case, it has been found that even if the above-mentioned fluttering falls, it can be determined that the body falls. Therefore, it can be said that it is possible to correctly detect the fall including flutter by selecting to detect the fall with the standard deviation of the standard deviation and setting the above parameters. In order to verify the effectiveness of the above parameters, as a result of verification of quiet fall, overcrowded (congested), stationary fall, and flapping fall, there is no false alarm when non-falling, and fall including flapping fall is about 5-10 Confirmed that the fall could be detected in seconds. That is, according to the verification result, when it is quiet, the occupation rate is as low as 12.5%, the standard deviation of the standard deviation is less than the threshold value for only a moment, and the range is small, so the object surface size hardly increases. Therefore, it is not determined to fall. During overcrowding (congestion), the occupancy rate exceeds 50% and the standard deviation of the standard deviation hardly falls below the threshold even at a very low speed of 4 m / min. Therefore, it is not erroneously determined to fall even in an overcrowded state. In addition, in the case of stationary falling, the difference from the moment of falling to the object surface size exceeding 200 mm is 3.5 seconds, and when the duration time is 1.5 seconds, the falling is detected in 5 seconds. In addition, in the case of falling down with fluttering, the difference from the moment of falling down to the object surface size exceeding 200 mm is 4.5 seconds, and when the duration time is 1.5 seconds, the falling is detected in 6 seconds. Actually, the local data was evaluated under the same conditions. Although the local data was very crowded, the standard deviation of the standard deviation was almost below the threshold and did not lead to false alarms.

Next, a fall detection algorithm of the passenger fall user fall detection device according to the basic example of the present invention will be described with reference to FIG.
The distance data for each measured angle is accumulated by the scan type distance sensor 7 composed of a laser scan sensor (step S1). Next, the measurement distance data is extracted (step S2), the standard deviation is calculated (step S3), and it is determined whether the standard deviation is equal to or less than the threshold (step S4). Here, the standard deviation is calculated from the data in the detection area for the past 1-2 seconds. Further, the measurement distance data is extracted (step S2), the standard deviation of the standard deviation is calculated (step S5), and it is determined whether or not the standard deviation of the standard deviation is equal to or less than the threshold value (step S6). Here, the standard deviation of the standard deviation is calculated from the standard deviations of the past 2 to 8 points. By the way, the calculation of the standard deviation (step S3) and the determination whether the standard deviation is equal to or less than the threshold (step S4) are not essential and may be omitted. If the standard deviation of the standard deviation is equal to or less than the threshold value, the surface size (length) of the object in that region is calculated (step S7). Then, as a result of the calculation, it is determined whether or not the surface size of the object is greater than or equal to a threshold value (step S8), and if the state equal to or greater than the threshold value lasts for a predetermined duration (step S9), it is determined that the object is in a falling state. (Step S10). Needless to say, when a fall is detected, the speed of the passenger conveyor is reduced and stopped. When a fall is detected, the alarm volume for warning may be increased as time passes. Further, when the scan angle of approximately 0 meters becomes a predetermined ratio or more, it may be determined that dirt is attached to the scan type distance sensor 7 (dirt detection function), and a warning may be issued.

  FIG. 8 is an explanatory diagram schematically showing the situation when the passenger conveyor is staying, FIG. 9 is a characteristic diagram showing the situation when the passenger conveyor is staying as a shadow area analysis result, and FIG. 10 is a shadow of the situation when the passenger conveyor is staying. Fig. 11 is a characteristic diagram showing the result of time integration of the area, Fig. 11 is a schematic diagram illustrating the situation when the passenger conveyor is overturned, and Fig. 12 is the shadow area analysis result showing the situation when the passenger conveyor is overturned. FIG. 13 is a characteristic diagram showing the result of time integration of the shadow area of the situation when the passenger conveyor falls over, and FIG. 14 shows the shadow area by the user fall detection device of the passenger conveyor in Embodiment 1 of the present invention. FIG. 15 is a flowchart for explaining the fall detection algorithm from the shadow area of the passenger fall detection device for passenger conveyors according to the first embodiment of the present invention. That. Here, the shadow area is an area of a region that hits the body portion.

  The characteristics of the present invention will be described with reference to FIGS. 8 to 13 while comparing the situation when the passenger conveyor is staying with the situation when the passenger rolls over. That is, after the surface of the fallen person is scanned using the scan type distance sensor 7, the area (shadow area) of the region hitting the trunk portion is extracted. When this shadow area is integrated over time, if the shadow continues for a certain amount of time, it is determined to fall. Since the person moves somewhat during the stay, the value does not increase when time integration is performed, and no false detection occurs. 8 and 9 show the situation at the time of stay and the shadow area analysis result, and FIG. 10 shows the shadow area time integration result. The circles in FIG. 9 are the user's feet captured by the scan-type distance sensor 7, and the squares are symbols indicating shadow areas. 11 and 12 show the situation and the shadow area analysis result when the flipping flaps, and FIG. 13 shows the shadow area time integration result. The meanings of the symbols in FIG. 12 are the same as those in FIG. Here, for example, if the value of the shadow area time integration, that is, the value obtained by time integration of the area of the region corresponding to the body portion is determined with a threshold value of 150 (square centimeter × second), at the time of residence, as shown in FIG. (Square centimeters x seconds) or less, so there is no false alarm. When falling over, as shown in FIG. 13, 150 (square centimeters x second) or more continues for about 10 seconds. .

  FIG. 14 is an explanatory diagram showing a state in which the fall state is analyzed by time integration of the shadow area by the passenger fall detection device for passenger conveyors according to the first embodiment of the present invention. In the figure, a scan type distance sensor 7 is used to laser scan the surface of the fallen person. A monitoring memory is configured in a two-dimensional array in the monitoring area. After scanning, shade the area that hits the body. The lines detected by the scan type distance sensor 7 are shown as bold lines. At this time, for example, four shadow areas are defined as a two-dimensional array in which the lines detected by the sensor enter and behind the scanning distance sensor 7 (upward in FIG. 14). This amount is determined by the size of one square of the two-dimensional array. Here, one square is 10 × 10 cm, and the shadow area is uniformly 40 cm backward. If it is no longer a shadow area in each cell (a state in which there is nothing), it is immediately set to zero. The first detected shadow area is set to 1. Also, the number of shadow regions that are continuous in time is increased. Here, the numerical value is updated every 0.1 second. The number of values equal to or greater than a predetermined threshold and the numerical value thereof are integrated, that is, the shadow area × time is calculated, and if it is equal to or greater than the predetermined value, an alarm is issued.

FIG. 15 is a flowchart for explaining a fall detection algorithm from a shadow area of a passenger fall detection device for a passenger conveyor according to the first embodiment of the present invention.
The distance is measured by the scan type distance sensor 7 formed of a laser scan sensor (step S11). Next, an object line in the monitoring area is extracted (step S12). Then, a shadow region is extracted behind the object line (seen from the sensor) (step S13). Here, the monitoring memory is constituted by a two-dimensional array, and the state where there is no shadow region is 0, and the first detected shadow region is 1. Then, in the next step S14, if the shadow area detected this time is also a shadow area in the previous time, the corresponding location in the monitoring memory is incremented by one. If the shadow area detected this time is not a shadow area last time, the corresponding location in the monitoring memory is set to 1. For the part that is not a shadow area this time, the corresponding part in the monitoring memory is set to zero. Next, the number in which the numerical value in the monitoring memory exceeds a predetermined value is counted to determine whether it is a predetermined threshold value (step S15). Here, if it is determined to be a predetermined threshold, an alarm is issued.

1 Passenger conveyor 2 First floor entrance / exit
3 Second floor entrance / exit (stop)
4 Floor board for getting on and off 5 Comb board 6 Passage for approach 7 Scan type distance sensor (laser scan sensor)
7a Laser beam scanning range 8 Processing device

Claims (3)

Passenger conveyors,
Floor boards for boarding / exiting, which are respectively installed at the entrances and exits of the passenger conveyor, and provided with a comb board on the step side of the passenger conveyor,
An approach passage for a passenger to approach on the side of the anti-comb plate of the floor board for getting on and off,
A scanning type distance sensor that is installed in the vicinity of the entrance / exit of the passenger conveyor, and that emits a laser beam in a horizontal plane in a scanning range including the floor board for getting on and off and an approach passage,
The scan type distance sensor scans the surface of a fallen person and time-integrates the area of the area hitting the body part to determine a fallen state when the value falls above a certain level. Detection device. Passenger conveyors,
Floor boards for boarding / exiting, which are respectively installed at the entrances and exits of the passenger conveyor, and provided with a comb board on the step side of the passenger conveyor,
An approach passage for a passenger to approach on the side of the anti-comb plate of the floor board for getting on and off,
A scanning type distance sensor that is installed in the vicinity of the entrance / exit of the entrance / exit of each of the passenger conveyors, and that emits a laser beam in a horizontal plane in a scanning range including the entrance / exit floor board and the approach passage,
The scan type distance sensor scans the surface of a fallen person and time-integrates the area of the area hitting the body part to determine a fallen state when the value falls above a certain level. Detection device.   The scan type distance sensor has a threshold value of about 150 (square centimeter x second) that is the time integral of the area of the area that hits the fuselage part, so that the situation at the time of dwelling and the situation at the time of falling over can be distinguished. The user fall detection device for a passenger conveyor according to claim 1 or 2, characterized in that:

Images