JP7642401B2 - Surroundings monitoring device for work vehicle and work vehicle - Google Patents

Description

The present invention relates to a surrounding monitoring device for a work vehicle.

Work vehicles such as forklifts move at slower speeds than regular vehicles, and they frequently switch between forward and reverse, and also make sudden turns, so their behavior is significantly different from that of regular vehicles. For this reason, it is difficult to transfer safety technology from regular vehicles directly to work vehicles.

Technology has been proposed to prevent accidents involving work vehicles. Patent Document 1 discloses technology that attaches RFID tags to people (workers) and objects that may pass near a work vehicle, and prevents contact based on the RFID tags.

JP 2009-1388 A

Conventional technology only detects people and objects equipped with RFID tags, so it is not possible to prevent contact with people who do not have RFID tags.

In addition, because the alarm is issued based solely on the relationship between the distance to the detected person or object and the vehicle speed, people who are simply walking along the aisle and therefore pose a low risk of contact are also subject to the warning. If the alarm continues to be issued in non-dangerous situations, its credibility will decrease, and work vehicle operators will become accustomed to such situations, making the alarm less useful.

The present invention has been made in consideration of these problems, and one exemplary purpose of one aspect of the present invention is to provide a perimeter monitoring device that can issue more appropriate warnings.

One aspect of the present disclosure relates to a periphery monitoring device for a work vehicle. The periphery monitoring device includes a person detection means for detecting people present in the vicinity of the work vehicle, a risk determination means for calculating a future predicted trajectory of the work vehicle and determining a risk level based on the relationship between the future predicted trajectory of the work vehicle and the detected people, and an alarm means for issuing an alarm in a manner corresponding to the risk level.

According to this aspect, by calculating a future predicted trajectory, the possibility of contact with a person can be estimated more accurately, and the degree of danger, which is an indicator of the possibility of contact, can be estimated more accurately, making it possible to issue an appropriate warning.

The risk determination means may calculate a future predicted trajectory of the detected person, and determine the risk based on the future predicted trajectory of the work vehicle and the future predicted trajectory of the detected person. By predicting the movement of the person as well, the risk can be evaluated with higher accuracy, enabling appropriate warnings to be issued.

The risk determination means may determine the risk based on at least one of the time and distance until contact with a detected person. Calculating the time until contact can reduce the number of warnings in situations where a person has ample time to escape. Calculating the risk based on distance can also provide an appropriate warning in cases where a work vehicle is stopped but is close to a person.

The risk determination means may determine the risk based on the distance between the work vehicle's future predicted trajectory and the detected person. Even if a person is outside the predicted passing range of the work vehicle, there is a possibility of contact due to a change in the steering angle, and this possibility increases the shorter the future predicted trajectory distance of the work vehicle. According to this aspect, it is possible to incorporate such potential dangers.

The blind spot area may be defined in advance. The risk determination means may determine the risk level based on the positional relationship between the detected person and the blind spot area. Work vehicles have their own unique blind spots. If a person is in such a blind spot, it will take a while for the operator to notice him or her. By including in the judgment factors whether or not the person is in a blind spot area, a more accurate risk assessment can be made.

The risk determination means may determine the risk based on the loading and unloading conditions. The loading and unloading conditions affect tipping and braking distance. Therefore, by including the loading and unloading conditions as a judgment factor, a more accurate assessment of the risk is possible.

The danger level determination means may determine the danger level based on the height at which the detected person is located. If the person is at a different level from the work vehicle, the danger level can be said to be low, so by including height as a determination factor, a more accurate assessment of the danger level can be made.

The warning means may intervene in braking based on the level of danger. In high-risk situations, braking assistance can be provided to increase safety.

In addition, any combination of the above components, or mutual substitution of components or expressions between methods, devices, systems, etc., are also valid aspects of the present invention.

The present invention makes it possible to give appropriate warnings.

1 is a perspective view showing the appearance of a forklift, which is one embodiment of a work vehicle. FIG. 2 is a block diagram of the periphery monitoring device. FIG. 13 is a diagram illustrating a future predicted trajectory. FIG. 13 is a diagram for explaining determination of a risk level based on a collision prediction time. 5A and 5B are diagrams for explaining a more advanced determination of the risk level based on the collision prediction time. FIG. 13 is a diagram for explaining the degree of risk based on the distance from a future predicted trajectory. FIG. 13 is a diagram illustrating a blind spot risk.

Preferred embodiments are described below with reference to the drawings. The same or equivalent components, parts, and processes shown in each drawing are given the same reference numerals, and duplicate descriptions are omitted where appropriate. Furthermore, the embodiments are illustrative and do not limit the invention, and all of the features and combinations thereof described in the embodiments are not necessarily essential to the invention.

Figure 1 is a perspective view showing the exterior of a forklift, which is one type of work vehicle. The forklift 600 includes a vehicle body (chassis) 602, forks 604L, 604R, a movable part (lifting body or lift) 606, a mast 608, and wheels 610, 612. The mast 608 is provided in front of the vehicle body 602. The movable part 606 is driven by a power source such as a hydraulic actuator (not shown in Figure 1) and moves up and down along the mast 608. Forks 604L, 604R for supporting cargo are attached to the movable part 606. The forklift 600 transports a pallet with the two forks 604L, 604R inserted into holes (fork pockets) in the pallet (not shown). Some forklift models have movable part 606 that can slide left and right.

The forklift 600 is equipped with a perimeter monitoring device to prevent contact with people in the vicinity. Figure 2 is a block diagram of the perimeter monitoring device 100. The perimeter monitoring device 100 is equipped with a person detection means 110, a danger level determination means 120, and an alarm means 130.

The human detection means 110 detects people in the vicinity of the forklift 600. For example, the human detection means 110 includes a sensor 112 and a processor 114. In this embodiment, as shown in FIG. 1, the sensor 112 is mounted so as to capture images of the rear of the vehicle body. This is because the forklift 600 often backs up, making it difficult for the operator (driver) to see the rear, and therefore the rear is more dangerous than the front.

For example, the sensor 112 is a stereo camera that can detect the position of a subject. The sensor 112 may be a LiDAR (Light Detection and Ranging) sensor, or a combination of a monocular camera and a distance sensor. The processor 114 analyzes the image from the sensor 112 to detect a person and output position information S1.

In addition to the person position information S1 from the person detection means 110, the risk determination means 120 also receives information S2 about the forklift 600 (hereinafter referred to as vehicle information). The vehicle information S2 may include (i) information about driving, such as vehicle speed and steering angle, (ii) information about loading and unloading, such as the weight of the cargo and the height of the forks, and (iii) information about the structure of the vehicle, such as vehicle width, overall length, and vehicle weight.

The risk determination means 120 determines a risk level S3 based on the person's position information S1 and the vehicle information S2. The risk level S3 is an index that correlates with the possibility of contact between the vehicle body and a person. The risk determination means 120 may be implemented in the same hardware as the arithmetic processing device 114, such as a microcomputer or CPU (Central Processing Unit), or may be implemented in separate hardware.

The warning means 130 issues a warning in a manner corresponding to the danger level S3 calculated by the danger level determination means 120. The warning means 130 is not particularly limited, but may include a visual warning means 132 that appeals to the operator's sense of sight. Examples of the visual warning means 132 include a display device (monitor) and a warning light, and a warning according to the danger level S3 can be issued by color, brightness, blinking, type of figure to be displayed, etc. For example, blue may be displayed when it is safe, yellow when the possibility of collision is low, and red when the possibility of collision is high.

The warning means 130 may include an auditory warning means 134 that appeals to the operator's sense of hearing. The auditory warning means 134 may include a simple buzzer or a voice reproduction means capable of more advanced voice guidance. In this case, a warning according to the danger level S3 can be issued by the type of sound, volume, and voice guidance.

The warning means 130 may include a physical warning means 136 that appeals to the operator's physical senses (such as touch) through vibration. For example, the operator may be notified by vibrating the steering wheel, levers, or seat to a degree that does not interfere with driving. The operator can be notified of the danger level S3 through the period and amplitude of the vibration.

The warning means 130 may include a braking assistance means 138. The braking assistance means 138 intervenes in the braking (brake) operation of the vehicle body according to the risk level S3, within a range that does not impair the stability of the vehicle body. In this specification, braking assistance is treated as one aspect of a warning.

The alarm means 130 may be one of the means 132, 134, 136, 138 illustrated herein, or may include any combination thereof, or may include others.

For example, when the risk level S3 is low, an alarm may be issued by one of the means 132, 134, or 136, and when the risk level S3 reaches its highest level, the brake assist means 138 may be activated.

Next, we will explain how the risk determination means 120 determines the risk level.

The risk determination means 120 calculates the future predicted trajectory of the forklift 600 based on the vehicle information S2. Then, it calculates the risk based on the relationship between the future predicted trajectory of the forklift 600 and the position of the person indicated by the person position information S1.

FIG. 3 is a diagram for explaining the future predicted trajectory. FIG. 3 shows a plan view of the body 602 of the forklift 600 seen from above. The future predicted trajectory 900 is a trajectory predicted to pass the reference point 603 of the body 602, and can be calculated based on dynamic information such as the current steering angle and shift lever position, and design values such as the dimensions of the body 602. Here, the reference point 603 is assumed to be the center of the trailing edge 602b of the body 602. The vehicle body passing predicted range 902 is a range through which the outline of the vehicle body 602 may pass in the future, and is hatched. The vehicle body passing predicted range 902 may have a width _W0 at the trailing edge 602b of the vehicle body 602 based on the future predicted trajectory 900, and the width W may gradually increase as it moves away from the trailing edge 602b.

Specifically, the vehicle body passing prediction range 902 is set to a range through which the vehicle body may pass when the vehicle is steered within a steering angle range and a steering operation speed range that can be operated without risk of rollover while taking dynamic stability into consideration in the current vehicle state. Here, the steering angle range and the steering operation speed range may be expressed in terms of the tire angle and the tire angular velocity, or may be expressed in terms of the operation angle and the operation angular velocity of the steering operation unit (steering wheel). In this embodiment, numerical values are given as examples of the tire angle and the tire angular velocity.

The current vehicle state is, for example, the current vehicle speed, the load of the luggage, and the lift height (and further the center of gravity and position of the vehicle). For example, if the current vehicle speed is 5 km/h and there is no load (no luggage), the vehicle body passing prediction range 902 is determined as ±10 degrees for the steering angle (tire angle) range and ±10°/s for the steering operation speed range (tire angular velocity), taking into account dynamic stability. Also, if the vehicle speed is 3 km/h, there is a luggage of 1000 kg, and the lift height is 3 m, the vehicle body passing prediction range 902 is determined as ±2 degrees for the steering angle range and ±5°/s for the steering operation speed range, taking into account dynamic stability. It is also possible to adopt the idea described in JP 2017-81662 A regarding the determination of the steering angle range and steering operation speed range that can be operated in the current vehicle state. In addition to the vehicle body passing prediction range 902 described above, the area into which a vehicle traveling on the side edge of the vehicle body passing prediction range 902 will enter if it rolls over to the outside may be set as a vehicle body overturning prediction area, and a risk level may be calculated and an alarm may be issued if a person is detected in the vehicle body overturning prediction area.

The risk determination means 120 can determine the final risk level S3 by quantifying the risk for each of several items and factors and integrating them. The items are explained below.

(1) Risk of collision based on predicted time R1
4 is a diagram for explaining the determination of the risk level R1 based on the collision prediction time. The risk level determination means 120 calculates the predicted time until a collision with a person 910 based on the future prediction trajectory 900, and the shorter the predicted time, the higher the risk level R1 based on the collision prediction time is set.

The simplest way is to calculate the time until the rear edge 602b of the vehicle body 602 comes into contact with the person 910, assuming that the person 910 is stationary. There are no particular limitations on the method of calculating the collision time.

For example, if the person 910 is included in the vehicle body passage prediction range 902, a perpendicular line may be dropped from the person 910 to the future prediction trajectory 900 to determine the intersection 912, and the collision time may be calculated from the length along the future prediction trajectory 900 from the reference point 603 to the intersection 912 and the vehicle speed. When the vehicle body 602 is stopped, the vehicle speed is zero, so the risk level R1 based on the collision prediction time is evaluated as low. Also, if the person 910 is not included in the vehicle body passage prediction range 902, the risk level R1 based on the collision prediction time is evaluated as low.

Alternatively, when the person 910 is included in the predicted vehicle passage range 902, a straight line 602b' on which the trailing edge 602b of the vehicle 602 may exist at each time point is calculated (Figure 4 shows straight lines at 0.5 seconds, 1 second, 1.5 seconds, etc.), and the predicted time until the person 910 comes into contact with this straight line 602b' is calculated to evaluate the risk level R1.

This risk level R1 is not based on a simple straight line between the vehicle body 602 and the person 910, so that a person 910 who is at low risk of collision can be prevented from being the subject of an alert.

Figures 5(a) and (b) are diagrams explaining how to determine a higher level of risk R1a based on the predicted collision time. The risk determination means 120 predicts the movement of the person 910 and calculates the future predicted trajectory 914 of the person 910. Then, taking into account both the future predicted trajectory 900 or the predicted vehicle passage range 902 of the vehicle body 602 and the future predicted trajectory 914 of the person 910, the collision time is predicted and the risk R1a is evaluated.

As shown in FIG. 5(a), when the future predicted trajectory 914a of the person 910 faces the direction of the vehicle body 602, the collision prediction time is shorter, and the risk level R1a is higher. Also, when the future predicted trajectory 914b of the person 910 faces in a direction away from the vehicle body passage prediction range 902, the possibility of a collision is lower, and the risk level R1a is lower.

In addition, at the risk level R1 in FIG. 4, the risk to a person 910 outside the vehicle body passage prediction range 902 is evaluated as low. On the other hand, at a high risk level R1a, as shown in FIG. 5(b), even if the person 910 is outside the vehicle body passage prediction range 902, if the future predicted trajectory 914c is heading toward the vehicle body passage prediction range 902, the possibility of a collision can be detected.

(2) Risk of collision based on predicted collision distance R2
The risk determination means 120 calculates a risk R2 based on the predicted collision distance. The predicted collision distance can be understood as the distance along the predicted future trajectory 900 between the reference point 603 and the intersection point 912 in FIG.

Alternatively, a trajectory 900' that passes through the person 910 and is perpendicular to the straight line 602b' at each time may be calculated, and the length along the trajectory 900' from the current vehicle trailing edge 602b to the person 910 may be calculated.

In calculating the risk level R2 based on the predicted collision distance, it may be considered whether the person 910 is located in an area that the forklift 600 cannot enter. For example, if image processing determines that the person 910 is on a sidewalk surrounded by a guard rail or on a sidewalk separated from the driving area of the forklift 600 by a step, the risk level R2 based on the predicted collision distance may be corrected to be lower.

(3) Risk level R3 based on distance from future predicted trajectory
6 is a diagram for explaining the risk level R3 based on the distance from the future predicted trajectory. The risk level R2 based on the above-mentioned collision prediction distance cannot take into account the risk of collision when the person 910 is located outside the vehicle body passage prediction range 902. Therefore, the risk level determination means 120 determines the risk level R3 based on the distance d between the future predicted trajectory 900 and the person 910. The distance d is the length of the perpendicular line drawn from the person 910 to the future predicted trajectory 900.

Even if the person 910 is located outside the vehicle passage prediction range 902, the shorter the distance d, the higher the possibility of a collision, and therefore the greater the risk R3. This risk R3 makes it possible to evaluate the risk of a collision due to factors such as a future change in the steering angle of the forklift 600.

(4) Blind spot danger level R4
7 is a diagram for explaining the blind spot risk level R4. A forklift has its own specific ranges (blind spots) that are difficult for the operator to see. Examples of blind spots include areas blocked by a head guard, a mast, or luggage, and areas directly behind the operator. A blind spot map is defined in which these blind spot areas 920 are divided into zones in advance. The risk level determination means 120 refers to the blind spot map and calculates the risk level R4 depending on which area 920 the person 910 is in.

If a person 910 is in the blind spot area 920, the operator will be late in discovering the person, and therefore the risk of collision will be higher than if the person 910 is outside the blind spot area 920. Therefore, by including whether or not the person 910 is in the blind spot area 920 as a judgment factor, it becomes possible to evaluate the degree of danger with higher accuracy.

(5) Fall risk level R5
The risk determination means 120 calculates the dynamic stability (instability) of the vehicle body 602 based on the load, lift height, traveling direction (tire angle and shift lever state), etc., and calculates a risk R5 based on the stability. The risk determination means 120 can calculate the risk of tipping over from known vehicle body information such as vehicle weight and center of gravity position, the loading situation, vehicle speed, and current values of front/rear/left/right acceleration.

Furthermore, the direction in which a person is likely to fall can be estimated using moment calculations at fall risk level R5, and fall risk level R5 can be determined based on the relationship between the person's position and the fall direction.

(6) Braking distance risk level R6
When the required braking distance is short, it is highly likely that a collision can be avoided by braking, whereas when the required braking distance is long, it becomes difficult to avoid a collision by braking.

The forklift may be equipped with a function to automatically control the degree of braking to prevent the load from shifting or tipping over. Specifically, if the load is heavy or tall, applying the brakes suddenly increases the possibility of tipping over. Therefore, when the possibility of tipping over is high, a control to reduce the braking may be incorporated. In this case, the braking distance will vary depending on the load, load height (lifting height), and traveling direction. Therefore, the required braking distance is calculated based on the load, lifting height, traveling direction, etc., and a braking distance risk level R6 according to the required braking distance is determined.

(7) Height of warning target danger level R7
The risk determination means 120 calculates a risk R7 based on the difference in height between the height of the person to be warned and the height of the vehicle. For example, if the platform on which the person is located is 1.5 m higher than the truck yard on which the forklift 600 runs, the possibility of a collision is low. Therefore, the risk determination means 120 can set the risk R7 high when the height of the vehicle and the height of the person are the same, and can set the risk R8 low when the difference in height is relatively large.

The above is an example of the risk calculated by the risk determination means 120. The risk determination means 120 calculates at least one or more of the risk levels R1 to R3 using the future predicted trajectory 900, and reflects these in the operation of the warning means 130.

The risk determination means 120 may calculate at least one of the risk levels R1 to R3, as well as at least one of the risk levels R4 to R7, and reflect these in the operation of the warning means 130.

The risk determination means 120 may calculate all risk levels R1 to R7, integrate them to calculate a total risk level, and control the warning means 130 based on the total risk level.

As an example, the risk determination means 120 may calculate an integrated risk by weighting and adding the risk levels R1 to R7.

As an example, since the risk levels R1 to R3 are in a parallel relationship, they can be weighted and added together, and the sum can be multiplied by the other risk levels R4 to R7 to calculate the overall risk level.

A complex function or neural network that defines the relationship between the risk levels R1 to R7 and the overall risk level may be defined, and the overall risk level may be calculated by inputting the risk levels R1 to R7 into this function or neural network.

The above is the configuration of the surroundings monitoring device 100. With this surroundings monitoring device 100, by calculating a future predicted trajectory, it is possible to more accurately estimate the possibility of contact with a person, and it is possible to issue an appropriate warning.

In addition, by providing gradual warnings according to the level of danger, it is expected that operators will be prevented from becoming accustomed to the warnings.

The individual benefits of the above risk levels R1 to R7 are explained below.

(1) Risk of collision based on predicted time R1
By calculating the risk level R1, it is possible to reduce the number of warnings or to lighten the level of warning in situations where the predicted time until collision is long and avoidance is fully possible, thereby preventing the operator from becoming accustomed to warnings.

(2) Risk of collision based on predicted collision distance R2
With only the risk level R1, it may not be possible to correctly evaluate the risk of collision when the vehicle is stopped. For example, even if the vehicle is stopped, if there is a person directly behind the vehicle, a collision will occur immediately after the vehicle starts to move, making it difficult to avoid. Therefore, by calculating the risk level R2, it is possible to evaluate the risk of collision immediately after a stopped vehicle starts to move.

(3) Risk level R3 based on distance from future predicted trajectory
By calculating the risk level R3, it is possible to take into consideration the change in the possibility of a collision accompanying a future change in the steering angle, although the possibility of a collision is low in the current vehicle state.

(4) Blind spot danger level R4
Blind spots are potential danger factors. Calculating the danger level R4 can prevent missed warnings. By increasing the danger level for people in positions that are easily overlooked by the operator, it becomes possible to warn them while there is still distance and time to do so.

(5) Fall risk level R5
By calculating the rollover risk R5 taking into account the dynamic stability of the vehicle body, the risk in the same situation can be evaluated more accurately than if it is not calculated.

(6) Braking distance risk level R6
Calculating the braking distance risk R6 enables a warning to be issued taking into consideration the distance and time allowed for avoidance.

(7) Height of warning target danger level R7
By calculating the height risk level R7, it is possible to eliminate warnings when the possibility of a collision is extremely low, or to reduce the level of the warning, thereby preventing the operator from becoming accustomed to the warnings.

The present invention has been described above based on examples. The present invention is not limited to the above-described embodiment, and it will be understood by those skilled in the art that various design changes and modifications are possible, and that such modifications are also within the scope of the present invention. These modifications are described below.

(Variation 1)
In the embodiment, the human detection means 110 monitors the rear and issues a warning when reversing, but instead, or in addition, the human detection means 110 may monitor the front of the vehicle body and issue a warning when moving forward.

(Variation 2)
In the embodiment, a forklift has been described as an example of a work vehicle, but the application of the present invention is not limited to this, and it can be applied to a hand lift (hand pallet), a fork loader (wheel loader with forks), etc.

100...periphery monitoring device, 110...human detection means, 112...sensor, 114...arithmetic processing device, 120...risk determination means, 130...alarm means, 132...visual alarm means, 134...auditory alarm means, 136...physical alarm means, 138...braking assistance means, 600...forklift, 602...vehicle body, 900...future predicted trajectory, 902...vehicle body passing predicted range, 910...person, 912...intersection, 914...future predicted trajectory, 920...blind spot area.

Claims (9)

A surrounding monitoring device for a work vehicle,
A person detection means for detecting a person present around the work vehicle;
a risk determination means for calculating a future predicted trajectory of the work vehicle, setting a vehicle body passing prediction range in which the work vehicle may pass in the future as a range that gradually widens with distance from the vehicle body based on the future predicted trajectory, and determining a risk level based on a relationship between the vehicle body passing prediction range and the detected person;
a warning means for issuing a warning in a manner corresponding to the degree of danger;
Equipped with
the risk determination means calculates a future predicted trajectory of the detected person, and determines the risk based on a time until contact with the detected person calculated from the future predicted trajectory of the work vehicle and the future predicted trajectory of the detected person ,
A surroundings monitoring device characterized in that the danger level determining means determines the danger level based on the difference in elevation between the height at which the detected person is located and the height at which the work vehicle is located. A surrounding monitoring device for a work vehicle,
A person detection means for detecting a person present around the work vehicle;
a risk determination means for calculating a future predicted trajectory of the work vehicle, setting a vehicle body passing prediction range through which the work vehicle may pass in the future based on the future predicted trajectory, setting an area into which the work vehicle may enter if it rolls over outward while traveling along a side end of the vehicle body passing prediction range as a vehicle body overturning prediction area, and determining a risk level based on a relationship between the vehicle body passing prediction range and the vehicle body overturning prediction area and the detected person;
a warning means for issuing a warning in a manner corresponding to the degree of danger;
A periphery monitoring device comprising: The surroundings monitoring device according to claim 1 or 2, characterized in that the risk determination means determines the risk based on a future predicted trajectory of the work vehicle and a distance from the detected person. The surroundings monitoring device according to any one of claims 1 to 3, characterized in that the blind spot area is defined in advance, and the danger level determination means determines the danger level based on the positional relationship between the detected person and the blind spot area. The surroundings monitoring device according to any one of claims 1 to 4, characterized in that the risk determination means determines the risk based on the loading situation. 3. The surroundings monitoring device according to claim 2 , wherein the danger level determining means determines the danger level based on a difference in elevation between the height at which the detected person is located and the height at which the work vehicle is located. A surrounding monitoring device for a work vehicle,
A person detection means for detecting a person present around the work vehicle;
a risk determination means for calculating a future predicted trajectory of the work vehicle and determining a risk level based on a relationship between the future predicted trajectory of the work vehicle and the detected person;
a warning means for issuing a warning in a manner corresponding to the degree of danger;
Equipped with
A surroundings monitoring device characterized in that the danger level determining means determines the danger level based on the difference in elevation between the height at which the detected person is located and the height at which the work vehicle is located. The surroundings monitoring device according to any one of claims 1 to 7, characterized in that the warning means intervenes in braking based on the degree of danger. A work vehicle equipped with a surroundings monitoring device according to any one of claims 1 to 8.

Images