Overturning prevention system

Abstract

To provide an overturning prevention system capable of preventing a vehicle body from overturning depending on the tilting degree of the vehicle body.SOLUTION: An overturning prevention system comprises a vehicle body 1 capable of driving straight forward and backward, a cargo handling device 2 including a fork, multiple wheels installed at an interval in the front-back direction and in the left-right direction, a forklift F with wheel load sensors 11A-11E to detect the loads acting on each of the multiple wheels, and a control device 17 to control the operations of the vehicle body 1 and of the cargo handling device 2. The control device 17 estimates the tilting degree of the vehicle body 1 using a machine learning based on the detected loads by the wheel load sensors 11A-11F and restricts the operations of the vehicle body 1 and the cargo handling device 2 based on the estimated result of the tilting degree. The control device 17 also restricts the operations of the vehicle body 1 and the cargo handling device 2 using the machine learning based on the detected loads by the wheel load sensors 11A-11E.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an anti-tip system for preventing the body of a forklift from tipping over. [Background technology]

[0002] To prevent the vehicle body from tipping over (including rolling over), a forklift is known that is configured to uniformly limit the traveling speed of the vehicle body when the fork lift height (height from the road surface to the fork) is equal to or greater than a predetermined height.

[0003] Furthermore, Patent Document 1 discloses a forklift that uses machine learning to restrict its driving and turning operations to prevent tipping over. Specifically, the forklift in Patent Document 1 acquires a tipping probability score indicating the likelihood of tipping over based on slippage data related to the degree of slippage of the forklift relative to the floor surface, balance data related to the balance of the forklift, and a learning model generated by machine learning, and restricts the drive wheel turning mechanism and drive wheel drive mechanism by reducing their drive speed or stopping them based on the tipping probability score. The slippage data includes the difference in rotational speed between the drive wheels and the driven wheels, and the balance data includes the type of forklift, the weight of the forklift, the position of the forks, the weight of the load being lifted or lowered by the forks, etc.

[0004] However, the configuration of Patent Document 1 does not restrict the operation of the forklift truck according to the degree of tilt of the vehicle body, so there is a risk that it may not be possible to prevent tipping. Specifically, for example, when the vehicle body is traveling on a sloped or ridged road surface, a low rollover likelihood score may be acquired and control to prevent tipping may not be performed, even though the vehicle body is prone to tipping and the likelihood of tipping is high. Also, for example, when the center of gravity of the load supported by the forks is significantly offset to the right or left from the lateral center axis of the vehicle body, a low rollover likelihood score may be acquired and control to prevent tipping may not be performed, even though the vehicle body is prone to tipping and the likelihood of tipping is high. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 6969858 Summary of the Invention [Problem to be solved by the invention]

[0006] The present invention has been made in consideration of the above circumstances, and an object of the present invention is to provide a rollover prevention system that can prevent a vehicle from rolling over depending on the degree of inclination of the vehicle body. [Means for solving the problem]

[0007] In order to solve the above problem, the anti-tip system of the present invention comprises a forklift equipped with a vehicle body capable of traveling straight forward and backward, a loading device including a fork, a plurality of wheels spaced apart in the forward and backward and left and right directions, and a wheel load sensor that detects the load acting on each of the plurality of wheels, and a control device that controls the operation of the vehicle body and the loading device, wherein the control device uses machine learning to estimate the degree of inclination of the vehicle body based on at least the load detected by the wheel load sensor, and restricts the operation of at least one of the vehicle body and the loading device based on the estimated degree of inclination.

[0008] The overturn prevention system further includes a vehicle speed sensor that detects the traveling speed of the vehicle body, a fork sensor that detects the moving speed of the fork relative to the vehicle body, a steering angle sensor that detects the steering angle of at least one of the wheels, and a storage device that stores a tilt degree prediction model generated by supervised learning using as training data the load acting on each of the plurality of wheels, the traveling speed of the vehicle body, the moving speed of the fork relative to the vehicle body, the steering angle of at least one of the wheels, and a label related to the degree of tilt of the vehicle body, and it is preferable that the control device estimates the degree of tilt of the vehicle body based on the tilt degree prediction model stored in the storage device.

[0009] In addition, in order to solve the above-mentioned problems, the anti-tip system of the present invention comprises a forklift equipped with a vehicle body capable of traveling straight forward and backward, a loading device including a fork, a plurality of wheels spaced apart in the forward and backward and left and right directions, and a wheel load sensor that detects the load acting on each of the plurality of wheels, and a control device that controls the operation of the vehicle body and the loading device, wherein the control device uses machine learning to limit the operation of at least one of the vehicle body and the loading device based on at least the load detected by the wheel load sensor.

[0010] Furthermore, it is preferable that the control device performs reinforcement learning based on at least the load detected by the wheel load sensor, determines an operation of at least one of the vehicle body and the loading device that will increase the reward as the degree of inclination of the vehicle body decreases, and restricts the operation of at least one of the vehicle body and the loading device so as to perform that operation.

[0011] Preferably, the wheel load sensor is a sensor incorporated in a rolling bearing that rotatably supports the wheel.

[0012] Furthermore, it is preferable that the plurality of wheels include a right front wheel positioned to the right of the fork, a left front wheel positioned to the left of the fork, and a drive wheel positioned rearward of the fork, and that the wheel load sensor is incorporated into a rolling bearing that rotatably supports the right front wheel, a rolling bearing that rotatably supports the left front wheel, and a rolling bearing that rotatably supports the drive wheel.

[0013] Furthermore, it is preferable that the plurality of wheels further include an auxiliary wheel that is arranged rearward of the fork and adjacent to the drive wheel in the left-right direction, and that the wheel load sensor is further incorporated into a rolling bearing that rotatably supports the auxiliary wheel. [Effects of the Invention]

[0014] According to the present invention, it is possible to provide a tip-over prevention system that can prevent tip-over depending on the degree of inclination of the vehicle body. [Brief explanation of the drawings]

[0015] [Figure 1] 1A and 1B are external views showing a schematic configuration of a forklift according to an embodiment of the present invention, in which (A) is a plan view and (B) is a side view. [Figure 2] FIG. 2 is a block diagram showing a schematic configuration of the forklift according to the embodiment. [Figure 3]10 is a flowchart showing a procedure for preventing the vehicle body from tipping over. DETAILED DESCRIPTION OF THE INVENTION

[0016] An embodiment of the present invention will be described with reference to the drawings. Note that the front-rear direction X, left-right direction Y, and vertical direction Z indicated by arrows in the drawings are linear directions that are perpendicular to one another.

[0017] As shown in Figures 1(A) and 1(B), the tip-over prevention system according to this embodiment is incorporated into a forklift truck F. The forklift truck F is a reach forklift truck, and includes a vehicle body 1, a cargo handling device 2, and a plurality of wheels 3.

[0018] The vehicle body 1 is configured to be able to travel straight in the longitudinal direction X, and also be able to turn to the right or left by changing the steering angle of the drive wheels 3B, which will be described later. The vehicle body 1 has a straddle leg 1L arranged across the cargo handling device 2 in the transverse direction Y. The straddle leg 1L is a right leg 1L provided on the right side of the cargo handling device 2. R and a left leg 1L provided on the left side of the loading device 2. L The vehicle body 1 also includes a traveling device 1A and a steering device 1B (both of which are shown in FIG. 2), which will be described later.

[0019] The cargo handling device 2 is configured to be able to lift cargo (not shown) and includes a fork 2A and a mast 2B. The fork 2A has two claws 2A extending parallel to each other with a gap between them. R ,2A L The mast 2B is configured as a telescopic mast that can be extended and retracted in the vertical direction Z, and supports the forks 2A. The cargo handling device 2 also includes a lift device 2C and a shift device 2D (both of which are described later in FIG. 2).

[0020] The wheels 3 are spaced apart in the front-rear direction X and the left-right direction Y. The wheels 3 include a right front wheel 3A R and left front wheel 3A L , drive wheels 3B, and auxiliary wheels 3C. Right front wheel 3A R and left front wheel 3A L is a road wheel that mainly receives the load acting on the fork 2A, and is a driven wheel that rolls on the road surface as the vehicle body 1 travels. R Right leg 1L R The left front wheel 3A is located to the right of the fork 2A. L Left leg 1L L It is located to the left of fork 2A.

[0021] The drive wheels 3B are connected to a motor (not shown) that generates power for the vehicle body 1 to travel, and roll on the road surface by the power transmitted from the motor. The drive wheels 3B are disposed behind the forks 2A, and the front wheels 3A R ,3A L The driving wheels 3B are rear wheels disposed behind the vehicle. The driving wheels 3B are steerable wheels configured so that the steering angle can be changed.

[0022] The auxiliary wheels 3C are rear wheels that are arranged behind the forks 2A, similar to the drive wheels 3B, and are arranged adjacent to the drive wheels 3B in the left-right direction Y. The auxiliary wheels 3C are driven wheels that roll on the road surface as the vehicle body 1 travels. The auxiliary wheels 3C are also dual-wheel casters that are configured so that the rolling direction can be changed according to the traveling direction of the vehicle body 1, and are a pair of wheels 3C that are spaced apart from each other. R ,3C L It is composed of:

[0023] As shown in FIG. 2, the forklift F is also provided with wheel load sensors 11A to 11E, a vehicle speed sensor 12, a lift sensor 13, a shift sensor 14, a steering angle sensor 15, a storage device 16, and a control device 17.

[0024] The wheel load sensors 11A to 11E are sensors incorporated in the rolling bearings 4A to 4E (see FIG. 1A) that rotatably support the wheels 3, respectively, and detect the load (i.e., wheel load) acting on each of the plurality of wheels 3. The wheel load sensor 11A detects the load acting on the right front wheel 3A. R The right front wheel 3A is mounted in a rolling bearing 4A that rotatably supports the right front wheel 3A. R The wheel load sensor 11B detects the wheel load of the left front wheel 3A. L The left front wheel 3A is mounted in a rolling bearing 4B that rotatably supports the left front wheel 3A. L The wheel load sensor 11C is incorporated in the rolling bearing 4C that rotatably supports the driving wheel 3B, and detects the wheel load of the driving wheel 3B. The wheel load sensor 11D detects the wheel load of the driving wheel 3B. R The wheel 3C is mounted in a rolling bearing 4D that rotatably supports the wheel 3C. R The wheel load sensor 11E detects the wheel load of the wheel 3C. L The wheel 3C is mounted in a rolling bearing 4E that rotatably supports the wheel 3C. L Detects the wheel load.

[0025] The vehicle speed sensor 12 detects the traveling speed of the vehicle body 1. The vehicle speed sensor 12 is configured, for example, by a wheel speed sensor that detects the wheel speed, which is the peripheral speed of the wheels 3. It is preferable that the wheel speed sensor detects the wheel speed of the drive wheels 3B, which are less likely to slip on the road surface.

[0026] The lift sensor 13 and the shift sensor 14 are fork sensors that detect the moving speed of the fork 2A relative to the vehicle body 1. The lift sensor 13 detects the moving speed of the fork 2A in the vertical direction Z relative to the vehicle body 1. The lift sensor 13 is configured, for example, by a distance measuring sensor that detects the height from the road surface to the fork 2A.

[0027] The shift sensor 14 detects the horizontal movement speed of the fork 2A relative to the vehicle body 1. In this embodiment, the shift sensor 14 is a reach sensor that detects the movement speed of the fork 2A in the front-rear direction X relative to the vehicle body 1. The shift sensor 14 is, for example, configured by a position sensor provided in the shift device 2D.

[0028] The steering angle sensor 15 detects the steering angle of the driving wheels 3B, which are steered wheels. The steering angle sensor 15 is configured, for example, by a rotation angle sensor that detects the rotation angle of a steering wheel for changing the steering angle of the driving wheels 3B.

[0029] The storage device 16 stores a tilt degree prediction model generated by supervised learning. The tilt degree prediction model is a model for predicting the degree of tilt, generated by supervised learning using as training data (A) the load acting on each of the multiple wheels 3, (B) the traveling speed of the vehicle body 1, (C) the moving speed of the forks 2A in the vertical direction Z relative to the vehicle body 1, (D) the moving speed of the forks 2A in the horizontal direction relative to the vehicle body 1, (E) the steering angle of the drive wheels 3B, and (F) a label related to the degree of tilt of the vehicle body 1. That is, of the training data, (A) to (E) are independent variables, and (F) is a dependent variable.

[0030] The independent variables for generating the tilt degree prediction model can be information detected by the sensors 11A-11E, 12-15 during a test run of the forklift F. It is preferable that the forklift F has exactly the same configuration as when it was tested (i.e., when the model was generated), but some of the configuration may have been changed since the test run. The label related to the tilt degree of the vehicle body 1, which is the dependent variable, is a tipping probability score indicating the possibility of tipping, and is information appropriately input by a label assigner.

[0031] The control device 17 controls the operation of the vehicle body 1 and the cargo handling device 2 to prevent the vehicle body 1 from tipping over. Specifically, the control device 17 estimates the degree of tilt of the vehicle body 1 using machine learning based on the load detected by the wheel load sensors 11A to 11E, and restricts the operation of the vehicle body 1 and the cargo handling device 2 based on the estimated degree of tilt. In this case, the control device 17 restricts the operation of the vehicle body 1 and the cargo handling device 2 using machine learning based on the load detected by the wheel load sensors 11A to 11E.

[0032] The control device 17 of this embodiment estimates the current degree of inclination of the vehicle body 1 based on the load detected by the wheel load sensors 11A-11E, the traveling speed of the vehicle body 1, the moving speed of the fork 2A, and the steering angle of the drive wheel 3B detected by the respective sensors 12-15 at the time of detecting the load, and the inclination degree prediction model stored in the storage device 16. That is, the control device 17 derives a tipping possibility score that indicates the current degree of inclination of the vehicle body 1 by inputting the information detected by the respective sensors 11A-11E and 12-15 into the inclination degree prediction model.

[0033] In addition, the control device 17 of this embodiment performs reinforcement learning based on the load detected by the wheel load sensors 11A to 11E, the traveling speed of the vehicle body 1, the moving speed of the fork 2A, and the steering angle of the drive wheel 3B detected by each sensor 12 to 15, and determines the operation of the vehicle body 1 and the loading device 2 that will increase the reward, assuming that the reward will be reduced when the degree of inclination of the vehicle body 1 decreases, and restricts the operation of the vehicle body 1 and the loading device 2 to perform that operation.

[0034] That is, the control device 17 that performs reinforcement learning determines an action a (i.e., the operation of the vehicle body 1 and the cargo handling device 2) that transitions the current state s observed using the sensors 11A-11E, 12-15 to a new state s' in an environment formulated as a Markov decision process. At this time, a reward corresponding to the transition to state s' is generated, and the state s' after the transition and the reward depend on the current state s and the action a. The Markov decision process can be solved by a value iteration method or a policy iteration method, and the reinforcement learning algorithm that can be used is, for example, a Q-learning algorithm, which is a reinforcement learning version of the value iteration method.

[0035] Furthermore, the control device 17 determines which of the traveling device 1A, the steering device 1B, the lift device 2C, and the shift device 2D will be subject to operation restriction, and the amount of operation restriction during the reinforcement learning. That is, the control device 17 selects one or more devices from the traveling device 1A, the steering device 1B, the lift device 2C, and the shift device 2D, and controls the amount of operation of the selected device so that it does not exceed the amount of operation restriction. It is preferable that the operation to be restricted is one that has little impact on work efficiency.

[0036] The traveling device 1A is configured with a prime mover that rotates the drive wheels 3B, and rotating the drive wheels 3B causes the vehicle body 1 to travel. When the operation of the traveling device 1A is restricted, the traveling speed of the vehicle body 1 is suppressed, thereby preventing the vehicle body 1 from tipping over.

[0037] The steering device 1B changes the steering angle of the drive wheels 3B to turn the vehicle body 1 to the right or left. When the operation of the steering device 1B is restricted, the turning radius of the vehicle body 1 is prevented from being reduced, thereby preventing the vehicle body 1 from tipping over.

[0038] The lift device 2C moves the forks 2A in the vertical direction Z relative to the vehicle body 1. When the operation of the lift device 2C is restricted, the upward displacement of the center of gravity of the forklift F including the load is suppressed, thereby preventing the vehicle body 1 from tipping over.

[0039] The shift device 2D moves the forks 2A horizontally relative to the vehicle body 1. The shift device 2D of this embodiment is a reach device that moves the forks 2A in the front-rear direction X relative to the vehicle body 1. When the operation of the shift device 2D is restricted, forward displacement of the center of gravity of the forklift F including the load is suppressed, thereby preventing the vehicle body 1 from tipping over.

[0040] The flow of preventing the vehicle body 1 from tipping over will be described with reference to Fig. 3. The forklift F repeatedly performs the series of processes shown in Fig. 3 as needed while traveling. Note that it is assumed that the storage device 16 has previously stored therein the tilt degree prediction model.

[0041] First, the sensors 11A to 11E, 12 to 15 detect the load acting on each of the multiple wheels 3, the traveling speed of the vehicle body 1, the moving speed of the fork 2A in the up-down direction Z and the fore-and-aft direction X, and the steering angle of the drive wheel 3B, and the control device 17 acquires various information for predicting the degree of tilt (step S1).

[0042] Next, the control device 17 derives a tipping possibility score based on the information acquired in step S1 and the tilt degree prediction model stored in the storage device 16, and estimates the current tilt degree of the vehicle body 1 (step S2).

[0043] Next, based on the estimation result of the degree of tilt in step S2, the control device 17 determines whether or not to perform anti-overturn control to prevent overturning of the vehicle body 1 (step S3). Specifically, when the overturn possibility score is less than a predetermined value, the control device 17 determines not to perform anti-overturn control, and when the overturn possibility score is equal to or greater than the predetermined value, the control device 17 determines to perform anti-overturn control.

[0044] When the control device 17 determines that anti-tip control should be performed (step S3: YES), it performs reinforcement learning based on the information acquired in step S1 to determine the operation of the vehicle body 1 and the loading device 2 that will reduce the degree of inclination of the vehicle body 1 (step S4).

[0045] Then, the control device 17 restricts the operation of the vehicle body 1 and the cargo handling device 2 so that the operation determined in step S4 is performed (step S5). In step S5, the operation of one or both of the vehicle body 1 and the cargo handling device 2 is restricted.

[0046] In this embodiment, the following effects are obtained. (1) The control device 17 estimates the degree of tilt of the vehicle body 1 using machine learning based on the loads and other factors detected by the wheel load sensors 11A-11E, and restricts the operation of the vehicle body 1 and the cargo handling device 2 based on the estimated degree of tilt. The loads detected by the wheel load sensors 11A-11E reflect the road surface condition of the vehicle body 1 and the positional deviation of the load supported by the forks 2A. Therefore, with the above configuration, when the vehicle body 1 is traveling on a sloping or ridged road surface, or when the center of gravity of the load supported by the forks 2A is significantly offset to the right or left from the lateral center axis of the vehicle body 1, the degree of tilt of the vehicle body 1 can be estimated to appropriately determine whether there is a high possibility of tipping over. As a result, the operation of the vehicle body 1 and the cargo handling device 2 is restricted based on the appropriate determination result of the possibility of tipping over, thereby preventing tipping over according to the degree of tipping over of the vehicle body.

[0047] Furthermore, according to the configuration (1) above, the control device 17 estimates the degree of inclination of the vehicle body 1 using machine learning, and therefore, by advancing the machine learning (i.e., by generating a model for predicting the degree of inclination using a large amount of training data), the accuracy of estimating the degree of inclination of the vehicle body 1 can be improved.

[0048] (2) The control device 17 estimates the degree of inclination of the vehicle body 1 based on the load detected by the wheel load sensors 11A to 11E, the traveling speed of the vehicle body 1 at the time of detecting the load, the moving speed of the fork 2A in the vertical direction Z at the time of detecting the load, the moving speed of the fork 2A in the horizontal direction at the time of detecting the load, the steering angle of the drive wheel 3B at the time of detecting the load, and the inclination degree prediction model stored in the storage device 16. With this configuration, the degree of inclination of the vehicle body 1 can be easily estimated using the inclination degree prediction model generated by machine learning.

[0049] (3) The control device 17 uses machine learning to restrict the operation of the vehicle body 1 and the loading device 2 based on the loads and other information detected by the wheel load sensors 11A-11E. The loads detected by the wheel load sensors 11A-11E reflect the road surface condition of the vehicle body 1 and the positional deviation of the load supported by the forks 2A. Therefore, with the above configuration, when the vehicle body 1 is traveling on an inclined or ridged road surface, or when the center of gravity of the load supported by the forks 2A is significantly offset to the right or left from the lateral central axis of the vehicle body 1, the operation of the vehicle body 1 and the loading device 2 can be restricted to prevent the vehicle from tipping over. In this way, tipping over can be prevented depending on the degree of inclination of the vehicle body 1.

[0050] In addition, in a conventional configuration in which the traveling speed of the vehicle body is uniformly limited when the fork height is equal to or greater than a predetermined height, there is a problem that the traveling speed is limited as long as the fork height is equal to or greater than the predetermined height, even when the possibility of the vehicle body tipping over is low, which reduces work efficiency. Furthermore, there is a problem that the traveling speed is not limited as long as the fork height is less than the predetermined height, even when the possibility of the vehicle body tipping over is high. Therefore, there is a need for improving work efficiency and safety.

[0051] According to the configuration (3) above, the control device uses machine learning to restrict the operations of the vehicle body 1 and the cargo handling device 2. Therefore, machine learning can be used to derive operations of the vehicle body 1 and the cargo handling device 2 that can achieve both work efficiency and safety, and by restricting the operations of the vehicle body 1 and the cargo handling device 2 to perform these operations, work efficiency and safety can be improved.

[0052] (4) The control device 17 performs reinforcement learning based on the loads and the like detected by the wheel load sensors 11A to 11E, and determines the operation of the vehicle body 1 and the cargo handling device 2 that will increase the reward as the degree of inclination of the vehicle body 1 decreases, and restricts the operation of the vehicle body 1 and the cargo handling device 2 so that the forklift F performs the determined operation. With this configuration, it is expected that appropriate restrictions will be applied to prevent the forklift F from tipping over, even when the forklift F performs complex operations.

[0053] (5) The wheel load sensors 11A to 11E are sensors incorporated into the rolling bearings 4A to 4E that rotatably support the wheels 3. This configuration can save space compared to, for example, a case where a sensor provided on the axle is used to detect the load acting on the wheels.

[0054] (6) Wheel load sensors 11A to 11C are connected to the right front wheel 3A. R a rolling bearing 4A that rotatably supports the left front wheel 3A; L The right front wheel 3A is mounted on a rolling bearing 4B that rotatably supports the right front wheel 3A, and the rolling bearing 4C that rotatably supports the driving wheel 3B. R , left front wheel 3A L , and the load acting on the drive wheels 3B can be detected, so that the distribution of the load in the front-rear direction X and the left-right direction Y can be obtained.

[0055] (7) The wheel load sensors 11D, 11E are incorporated into the rolling bearings 4D, 4E that rotatably support the training wheel 3C. This configuration allows the load acting on the training wheel 3C to be further detected, making it possible to obtain a highly accurate load distribution.

[0056] The present invention is not limited to the above-described embodiment, and the above configurations can be modified. For example, the following modifications can be made, or the following modifications can be combined to make the present invention.

[0057] The control device may limit the operation of only one of the traveling device and the steering device as the operation of the vehicle body. In other words, the control device may limit the operation of at least one of the traveling device and the steering device as the operation of the vehicle body.

[0058] The control device may limit the operation of only one of the lift device and the shift device as the operation of the cargo handling device. In other words, the control device may limit the operation of at least one of the lift device and the shift device as the operation of the vehicle body.

[0059] The control device may limit the operation of only one of the vehicle body and the cargo handling device. In other words, the control device may limit the operation of at least one of the vehicle body and the cargo handling device.

[0060] The control device may use machine learning in a manner other than that described in the above embodiment as long as it can estimate the degree of tilt of the vehicle body based on the load detected by the wheel load sensor. That is, the control device may use machine learning to estimate the degree of tilt of the vehicle body based on at least the load detected by the wheel load sensor.

[0061] The control device may use machine learning in a manner other than that of the above-described embodiment as long as it can limit the operation of at least one of the vehicle body and the loading device based on the load detected by the wheel load sensor. That is, the control device may use machine learning to limit the operation of at least one of the vehicle body and the loading device based on at least the load detected by the wheel load sensor.

[0062] The forklift may further include a side shift device that moves the forks left and right as a shift device that moves the forks horizontally relative to the vehicle body. In this case, it is preferable that the shift sensor also detects the speed of the forks moving left and right relative to the vehicle body.

[0063] The forklift is not limited to a reach forklift truck, but may also be a counterbalance forklift truck or a three-way stacking truck. In a three-way stacking truck, where the forks do not move forward or backward relative to the vehicle body, the shift sensor may detect the speed of the forks moving left and right, rather than forward or backward.

[0064] The steering angle sensor may detect the steering angle of a wheel other than the drive wheels as long as it can detect the steering angle of the steering wheels. The fork sensor may be composed of at least one of a lift sensor and a shift sensor as long as it can detect the moving speed of the fork relative to the vehicle body.

[0065] The anti-tip system may include a control device and a storage device external to the forklift, and may be configured to limit the operation of the vehicle body and the loading / unloading device based on commands from the external device.

[0066] The fork configuration may be changed depending on the load handled by the forklift. For example, the forklift may have two prongs 2A. R ,2A L Instead of the fork 2A configured with three or more prongs, a fork configured with one cylindrical ram may be provided. [Explanation of symbols]

[0067] 1 Vehicle body 1A Running gear 1B Steering gear 2. Cargo handling equipment 2A Fork 2C Lifting device 2D shift device 3 wheels 3A R Right front wheel 3A L left front wheel 3B Drive Wheel 3C training wheels 4A~4D Rolling bearings 11A~11E Wheel load sensor 12 Vehicle speed sensor 13 Lift sensor (fork sensor) 14 Shift sensor (fork sensor) 15 Steering angle sensor 16 Storage device 17 Control device F Forklift X Anteroposterior direction Y left / right direction Z vertical direction

Claims

1. a forklift truck including a vehicle body capable of traveling straight forward and backward, a loading device including a fork, a plurality of wheels spaced apart in the forward and backward and left and right directions, and a wheel load sensor that detects the load acting on each of the plurality of wheels; a control device for controlling the operation of the vehicle body and the loading device, The control device estimates the degree of inclination of the vehicle body using machine learning based on at least the load detected by the wheel load sensor, and restricts the operation of at least one of the vehicle body and the loading device based on the estimated result of the degree of inclination. A fall prevention system characterized by:

2. a vehicle speed sensor for detecting a traveling speed of the vehicle body; a fork sensor that detects a moving speed of the fork relative to the vehicle body; a steering angle sensor that detects a steering angle of at least one of the wheels; a storage device that stores an inclination degree prediction model generated by supervised learning using as training data the load acting on each of the plurality of wheels, the traveling speed of the vehicle body, the moving speed of the fork relative to the vehicle body, the steering angle of at least one of the wheels, and a label related to the inclination degree of the vehicle body, The control device estimates the degree of inclination of the vehicle body based on the inclination degree prediction model stored in the storage device. The fall prevention system according to claim 1 .

3. a forklift truck including a vehicle body capable of traveling straight forward and backward, a loading device including a fork, a plurality of wheels spaced apart in the forward and backward and left and right directions, and a wheel load sensor that detects the load acting on each of the plurality of wheels; a control device for controlling the operation of the vehicle body and the loading device, The control device uses machine learning to limit the operation of at least one of the vehicle body and the loading device based on at least the load detected by the wheel load sensor. A fall prevention system characterized by:

4. The control device performs reinforcement learning based on at least the load detected by the wheel load sensor, and determines an operation of at least one of the vehicle body and the cargo handling device that will increase the reward as the degree of inclination of the vehicle body decreases, and restricts the operation of at least one of the vehicle body and the cargo handling device so as to perform the operation.

4. The fall prevention system according to claim 1, wherein the fall prevention system is a system for preventing a fall.

5. The wheel load sensor is a sensor incorporated in a rolling bearing that rotatably supports the wheel.

4. The fall prevention system according to claim 1, wherein the fall prevention system is a system for preventing a fall.

6. the plurality of wheels include a right front wheel disposed to the right of the fork, a left front wheel disposed to the left of the fork, and a drive wheel disposed rearward of the fork, The wheel load sensor is incorporated into a rolling bearing that rotatably supports the right front wheel, a rolling bearing that rotatably supports the left front wheel, and a rolling bearing that rotatably supports the drive wheel. The fall prevention system according to claim 5 .

7. The plurality of wheels further includes an auxiliary wheel that is disposed rearward of the fork and adjacent to the drive wheel in the left-right direction, The wheel load sensor is further incorporated in a rolling bearing that rotatably supports the auxiliary wheel. The fall prevention system according to claim 6 .

Images