JP7478092B2 - Cargo handling system - Google Patents

Description

The present invention relates to a cargo handling vehicle system that controls a cargo handling vehicle.

In recent years, labor shortages caused by a declining birthrate and aging population, and an increase in the number of logistics transactions due to the expansion of the e-commerce market have made it necessary to reduce the number of workers required in logistics warehouses and improve work efficiency. To solve this problem, the introduction of unmanned industrial vehicles that can operate unmanned is being promoted.

Loading vehicles such as forklifts are used to transport cargo within warehouses. In an unmanned transport system using loading vehicles, it is necessary to automatically perform a series of operations, such as the loading operation in which the tips of the loading parts (forks) are inserted into the opening of a pallet (loading platform) carrying the cargo and lifting it up, and the unloading operation in which the cargo loaded on the loading parts is moved to a designated position and unloaded by removing the loading parts.

In order to safely and efficiently automatically carry out cargo transportation work using cargo handling vehicles as described above, a driving control technology that enables autonomous control of the cargo handling vehicles is required. One driving control technology is a path following function that enables the cargo handling vehicle to drive autonomously along a target route. To build a path following function, a plant model of the vehicle to be controlled (a model that shows the operating characteristics of the vehicle to be controlled) is generally created based on the geometric conditions of the vehicle structure. After that, a feedback control model is built that minimizes the difference between the target route and the vehicle's own position, and the parameters of the plant model and feedback control model are tuned to achieve a highly accurate path following function.

Loading vehicles such as forklifts perform multiple tasks, not only driving and stopping, but also turning on the spot (spin turns), raising and lowering the loading and unloading device (a device that raises and lowers the loading components), and moving the loading components back and forth. In addition, loading vehicles perform driving operations depending on whether or not there is a load, and driving operations in which the loading components are positioned behind the direction of travel (reverse driving) to prevent the load from falling off while driving, so there are various driving modes for loading vehicles.

When running vehicles with the wide variety of driving modes described above, it is extremely difficult to construct a highly accurate plant model and feedback control model that can handle all of the tasks that need to be performed.

To address these issues, for example, the invention described in Patent Document 1 has been proposed. The industrial vehicle (forklift) described in Patent Document 1 is equipped with multiple driving modes according to the work and a plant model corresponding to the driving mode, and the forklift driver manually switches the driving mode according to the work being performed.

JP 2017-226514 A

In conventional technologies such as the invention described in Patent Document 1, a worker such as a forklift driver manually switches the driving mode depending on the work being performed, thereby switching the plant model. In an autonomously controlled loading and unloading vehicle, it is necessary to automatically perform tasks such as driving, loading, unloading, and turning on the spot, and for this reason, it is also necessary to automatically switch the plant model.

The present invention aims to provide a loading vehicle system that can automatically switch plant models.

The loading vehicle system according to the present invention includes a loading member on which the loading vehicle carries the load, a control command generating unit, a travel route generating unit, a nearest node calculating unit, a work determining unit, and a plant model switching unit. The control command generating unit controls the loading vehicle according to a plant model that indicates the operating characteristics of the loading vehicle when performing work. The travel route generating unit creates a travel route for the loading vehicle. The travel route is composed of a plurality of nodes that indicate positions through which the loading vehicle passes, and one or more links that connect two of the nodes. Each of the nodes is provided with a work ID that indicates the work performed by the loading vehicle. The nearest node calculating unit inputs the position of the loading vehicle and the travel route, and determines, as a current node, the node that is closest to the loading vehicle among the nodes that make up the travel route, and determines, as a target node, the node that the loading vehicle should pass through next after the current node among the nodes that make up the travel route. The work determining unit compares the work ID at the current node with the work ID at the target node. If the work ID at the current node and the work ID at the target node are different from each other, the plant model switching unit switches the plant model to the plant model corresponding to the work identified by the work ID at the target node.

The present invention provides a loading vehicle system that can automatically switch plant models.

1 is a diagram showing a loading vehicle controlled by a loading vehicle system according to a first embodiment of the present invention; 1 is a functional block diagram showing a configuration of a loading vehicle system according to a first embodiment. FIG. 2 is a diagram showing an outline of route information for loading and unloading vehicles. 11 is a diagram showing work performed by a loading vehicle for each work ID. FIG. 10 is a flowchart showing a flow of processing by an operation determination unit and a plant model switching unit. 13 is a flowchart showing a flow of another process of an operation determination unit and a plant model switching unit. FIG. 13 is a conceptual diagram showing a loading vehicle when switching the plant model. FIG. 4 is a diagram showing a transfer function of a traveling motor. 5 is a diagram illustrating an example of a relationship between a command value and a response value of a traveling motor. FIG. FIG. 4 is a diagram showing a transfer function of a steering motor. 5A and 5B are diagrams illustrating an example of the relationship between a command value and a response value of a steering motor. FIG. 1 is a diagram showing a loading vehicle traveling forward (rear-wheel drive). FIG. 1 is a diagram showing a loading vehicle that travels backward (front-wheel drive). FIG. 6 is a functional block diagram showing a configuration of a loading vehicle system according to a second embodiment of the present invention. FIG. 11 is a diagram showing a transfer function indicating the responsiveness of a steering motor in the second embodiment. 13 is a diagram showing feedforward control using a transfer function indicating the responsiveness of a steering motor obtained by a learning unit in the second embodiment. FIG.

The loading vehicle system according to the present invention can automatically switch the plant model for controlling the loading vehicle according to the work of the loading vehicle. For example, the loading vehicle system according to the present invention automatically acquires the work to be performed from information on the travel route, and automatically switches the plant model based on the acquired work to be performed. The loading vehicle system according to the present invention may also be configured to automatically generate data (learning data) for improving the accuracy of the plant model for each work.

A plant model is a model that shows the operating characteristics of a loading vehicle when it performs work, and is also called a vehicle model. The plant model is expressed by transfer functions and equations of motion, as described below.

The cargo handling vehicle system of the present invention automatically switches plant models according to the work to be performed, enabling highly accurate driving control and improving the transport operation efficiency of the cargo handling vehicle. In addition, because the plant model is switched for each work, restrictions on the paths along which the cargo handling vehicle travels can be reduced, which also leads to optimizing the layout of the warehouse where the cargo to be loaded onto the cargo handling vehicle is stored. Furthermore, if the cargo handling vehicle system of the present invention is configured to automatically generate data for each work to improve the accuracy of the plant model, a tuning-free control system can be constructed, thereby reducing the engineering costs at the time of introducing the system.

The following describes a cargo handling vehicle system according to an embodiment of the present invention. In the embodiment described below, the cargo handling vehicle controlled by the cargo handling vehicle system is a cargo handling vehicle capable of unmanned operation and autonomous control to perform tasks automatically, such as a forklift that travels within a warehouse.

In addition, in the drawings used in this specification, the same or corresponding components are given the same reference numerals, and repeated explanations of these components may be omitted.

The cargo handling vehicle system according to the first embodiment of the present invention will be described with reference to Figures 1 to 10.

First, we will explain the cargo handling vehicle controlled by the cargo handling vehicle system of this embodiment.

(Loading vehicle 100)
1 is a diagram showing a cargo handling vehicle 100 controlled by a cargo handling vehicle system according to the present embodiment. The cargo handling vehicle 100 is capable of autonomous control and transports cargo.

The loading vehicle 100 comprises a vehicle frame 101, a transfer device 102 that is mounted on the vehicle frame 101 so as to be able to rise and fall, and a loading member 103 that can move in and out of the transfer device 102 and that loads cargo. For example, the loading vehicle 100 is a forklift that travels within a warehouse, and the loading member 103 is a fork. In FIG. 1, the elevated transfer device 102 and loading member 103 are shown by dashed lines.

An external sensor 104 is installed on the loading vehicle 100 to acquire position information of objects around the loading vehicle 100. The external sensor 104 is installed, for example, on the upper part of the vehicle frame 101. The loading vehicle 100 can be installed with one or more external sensors 104. The external sensor 104 is, for example, a LiDAR (Light Detection and Ranging) device, and detects position information and shape information of objects around the loading vehicle 100 as a point cloud by changing the irradiation direction of a laser light at a predetermined angle (for example, at 0.5 degrees).

The cargo handling vehicle 100 is equipped with an acceleration sensor 105 that acquires the acceleration when the vehicle body is traveling. The acceleration sensor 105 can also acquire the angular velocity when the vehicle body of the cargo handling vehicle 100 turns. The acceleration sensor 105 is installed, for example, on the side of the vehicle frame 101. The cargo handling vehicle 100 can be equipped with one or more acceleration sensors 105.

In the loading vehicle 100, a load sensor 106 that acquires the mass of the load is installed on the loading member 103. The load sensor 106 is, for example, a pressure sensor. The loading vehicle 100 can be equipped with one or more load sensors 106. For example, the loading vehicle 100 may be equipped with multiple load sensors 106 arranged in the fore-and-aft direction of the loading member 103 and an inclination sensor that detects the inclination angle of the loading member 103, and may be configured to determine the vertical center of gravity position of the load with high accuracy.

The loading vehicle 100 is provided with an on-board controller 107 that performs calculations for autonomous control of the loading vehicle 100. The on-board controller 107 is provided, for example, on the upper surface of the vehicle frame 101. One or more on-board controllers 107 are provided in the loading vehicle 100. For example, the loading vehicle 100 may be provided with multiple on-board controllers 107, and each on-board controller 107 may perform different processes.

(Loading vehicle system 200)
Fig. 2 is a functional block diagram showing the configuration of a cargo handling vehicle system 200 according to this embodiment. The cargo handling vehicle system 200 controls the cargo handling vehicles 100 capable of autonomous control. The cargo handling vehicle system 200 can control one or a plurality of cargo handling vehicles 100. In Fig. 2, solid lines with arrows indicate data flow.

The loading vehicle system 200 comprises a traffic control unit 201 and an on-board controller 107 installed on the loading vehicle 100. The traffic control unit 201 comprises a communication device 203. The on-board controller 107 comprises a communication device 202. The traffic control unit 201 and the on-board controller 107 connect to each other and communicate via their respective communication devices 203, 202. The traffic control unit 201 and the on-board controller 107 can send and receive necessary data to each other using a wireless network or the like.

Furthermore, the cargo handling vehicle system 200 includes an external sensor 104, an acceleration sensor 105, and a load sensor 106 installed on the cargo handling vehicle 100. The on-board controller 107 is connected to the external sensor 104, the acceleration sensor 105, and the load sensor 106, and inputs information acquired by these sensors.

The on-board controller 107 includes a plant model switching unit 231, which will be described later, and automatically switches the plant model depending on the task to be performed.

The traffic control unit 201 and the on-board controller 107 can be configured as a computer equipped with a processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) for performing calculations to control the loading vehicle 100, and a memory for storing programs and data.

(Traffic Control Unit 201)
First, a description will be given of the traffic control unit 201. The traffic control unit 201 includes an operation management unit 204, a route map management unit 205, a travel route generation unit 206, a plant model management unit 232, and the above-mentioned communication device 203, and performs operation management of the cargo handling vehicle 100.

(Operation management unit 204)
The operation management unit 204 determines a task command to be executed by the cargo handling vehicle 100 and outputs it to the travel route generation unit 206. The operation management unit 204 stores in advance a plurality of task commands that determine which cargo is to be transported from where to where. The operation management unit 204 assigns these task commands to the plurality of cargo handling vehicles 100, thereby determining the task command to be executed by each of the cargo handling vehicles 100. For example, the operation management unit 204 can assign task commands to the cargo handling vehicles 100 according to the positions of the cargo handling vehicles 100.

(Route Map Management Unit 205)
The route map management unit 205 stores data of a two-dimensional map of the range in which the cargo handling vehicle 100 travels. Information (route information) about the routes on which the cargo handling vehicle 100 can travel is added to this two-dimensional map by the operator of the cargo handling vehicle system 200. The two-dimensional map is, for example, a map of the entire warehouse in which the cargo handling vehicle 100 travels. The route map management unit 205 can input point cloud data acquired by the external sensor 104 while the cargo handling vehicle 100 is traveling, and create a two-dimensional map in advance using SLAM (Simultaneous Localization and Mapping), for example.

Fig. 3 is a diagram showing an overview of the route information of the cargo handling vehicle 100. Fig. 3 shows, as an example, a route by which the cargo handling vehicle 100 approaches a shelf 307 and a pallet 304 on which a load is placed. A reference coordinate system 310 is set in the two-dimensional map stored by the route map management unit 205. In Fig. 3, the reference coordinate system 310 is represented by an _XG axis and a _YG axis.

The route information is composed of multiple travel nodes 301, 302 indicating the positions through which the loading vehicle 100 passes, and one or more links 303 connecting the two travel nodes 301, 302. One link 303 or multiple links 303 connected to each other constitute a segment 305. The route information is composed of one or more segments 305. The travel node located at the end of the segment 305 (the end of the route) is the end node 306.

Each of the traveling nodes 301, 302 has data such as the target position, target attitude, target speed, and turning radius R for the cargo handling vehicle 100 relative to the reference coordinate system 310. The target position, target attitude, and target speed are the position, attitude, and speed that the cargo handling vehicle 100 should take at each traveling node 301, 302. The turning radius R is the radius that the cargo handling vehicle 100 should take when turning at each traveling node 301, 302. These data are added to the two-dimensional map by the operator of the cargo handling vehicle system 200 as route information.

The attitude of the loading vehicle 100 refers to the yaw angle (azimuth angle) of the body of the loading vehicle 100 relative to the reference coordinate system 310.

The current position of the loading vehicle 100 is referred to as the self-position 311, and the current attitude of the loading vehicle 100 is referred to as the self-attitude 312. In Fig. 3, the self-attitude 312 of the loading vehicle 100 is represented by the angle of the longitudinal direction _YV of the loading vehicle 100 with respect to the _YG axis (equal to the angle of the lateral direction _XV of the loading vehicle 100 with respect to the _XG axis).

In addition, a work ID is preset in the route map management unit 205. The work ID indicates the work to be performed by the loading vehicle 100. A work ID is provided for each of the traveling nodes 301, 302 in the route information stored by the route map management unit 205. The work to be performed by the loading vehicle 100 at each of the traveling nodes 301, 302 is specified by the work ID. One traveling node can be provided with multiple work IDs.

Figure 4 is a diagram showing the work performed by the loading vehicle 100 for each work ID. The work performed by the loading vehicle 100 includes, as main tasks, for example, straight-line driving (driving with a turning radius R larger than a predetermined radius R0), turning driving (driving with a turning radius R equal to or smaller than radius R0), spin turns (turning on the spot without moving), and loading and unloading. The radius R0 can be set arbitrarily in advance, for example, 360 m. Furthermore, the work performed by the loading vehicle 100 includes, as subtasks, for example, the direction of movement, the direction of turning, and the type of loading and unloading (loading operation or unloading operation). In addition, the work performed by the loading vehicle 100 is specified by the work ID depending on whether the loading vehicle 100 is loaded with a load (presence or absence of a load).

Returning to Figure 2, we will continue explaining the traffic control unit 201.

(Drive route generation unit 206)
The travel route generating unit 206 inputs a task command to be executed by the cargo handling vehicle 100 from the operation management unit 204, and inputs route information from the route map management unit 205, to generate a travel route for the cargo handling vehicle 100 to execute the task command. The travel route generating unit 206 can generate a travel route for the cargo handling vehicle 100 using existing technology. The travel route generated by the travel route generating unit 206 is composed of a plurality of travel nodes 301, 302 including a terminal node 306, and one or more links 303 connecting the two travel nodes 301, 302, as shown in FIG. 3, for example. As described above, the travel nodes 301, 302 are provided with a work ID.

The driving route generation unit 206 transmits the created driving route of the loading/unloading vehicle 100 to the self-position and attitude calculation unit 214 of the on-board controller 107 via the communication device 203 of the traffic control unit 201 and the communication device 202 of the on-board controller 107.

The plant model management unit 232 will be explained later.

(In-vehicle controller 107)
Next, a description will be given of the on-board controller 107. The on-board controller 107 includes a self-position and attitude calculation unit 214, a nearest node calculation unit 215, a load calculation unit 216, a vehicle body behavior calculation unit 217, a control command generation unit 220, a steering motor drive unit 221, a travel motor drive unit 222, a brake drive unit 223, a transfer device and loading member motor drive unit 224, a work determination unit 230, a plant model switching unit 231, and the above-mentioned communication device 202, and performs calculations with these elements.

(Self-position and orientation calculation unit 214)
The self-position and attitude calculation unit 214 acquires the self-position 311 and self-attitude 312 ( FIG. 3 ) of the cargo handling vehicle 100 in the reference coordinate system 310, using point cloud information on objects around the cargo handling vehicle 100 acquired by the external sensor 104 and data on the two-dimensional map provided by the route map management unit 205. The self-position and attitude calculation unit 214 transmits the acquired self-position 311 and self-attitude 312 to the nearest node calculation unit 215. The self-position and attitude calculation unit 214 also transmits the travel route of the cargo handling vehicle 100 received from the traffic control unit 201 to the nearest node calculation unit 215.

The self-position/attitude calculation unit 214 may obtain the self-position 311 and self-attitude 312 of the cargo handling vehicle 100 using other means. For example, the self-position/attitude calculation unit 214 may use information such as the angular velocity of the cargo handling vehicle 100 when turning detected by the acceleration sensor 105, the rotational speed of the wheels of the cargo handling vehicle 100, and the radius of the wheels to more accurately determine the self-position 311 and self-attitude 312. The rotational speed of the wheels of the cargo handling vehicle 100 can be detected by an encoder installed on the cargo handling vehicle 100.

(Nearest node calculation unit 215)
The nearest node calculation unit 215 calculates the traveling node that is closest to the self-position 311 of the cargo handling vehicle 100 as the nearest node among the traveling nodes that configure the traveling route of the cargo handling vehicle 100. As described above, the traveling route of the cargo handling vehicle 100 is generated by the traveling route generation unit 206 of the traffic control unit 201. As described above, the self-position 311 of the cargo handling vehicle 100 is acquired by the self-position/attitude calculation unit 214.

The nearest node calculation unit 215 inputs the self-position 311 and travel route of the cargo handling vehicle 100, and finds the nearest node of the cargo handling vehicle 100 by calculating the distance difference between the self-position 311 and the travel node group that constitutes the travel route. The nearest node of the cargo handling vehicle 100 is called the "current node." In the example shown in FIG. 3, the travel node 301 is the current node.

Furthermore, among the traveling nodes that make up the traveling route of the loading vehicle 100, the traveling node (traveling node 302 in FIG. 3) that the loading vehicle 100 should pass through next after the current node (traveling node 301 in FIG. 3) is called the "target node". The target node is a traveling node (traveling node 302 in FIG. 3) other than the current node among the links connected to the current node (traveling node 301) and link 303 that exists in the traveling direction of the loading vehicle 100. The link 303 that connects the current node (traveling node 301) and the target node (traveling node 302) is called the "target link". In the target link (link 303), the direction from the current node (traveling node 301) to the target node (traveling node 302) is called the "target traveling direction".

The nearest node calculation unit 215 finds the current node 301, the target node 302, and the target link 303, and transmits information on the current node 301, the target node 302, and the target link 303, as well as information defined therefor, to the control command generation unit 220 and the work determination unit 230. The nearest node calculation unit 215 also transmits the work IDs provided for the current node 301 and the target node 302, and the self-position 311 of the loading vehicle 100 to the control command generation unit 220 and the work determination unit 230.

(Load Weight Calculation Unit 216)
The live load calculation unit 216 acquires the mass of the load using the load sensor 106. The live load calculation unit 216 transmits the acquired mass of the load to the task determination unit 230 and the control command generation unit 220.

(Vehicle body behavior calculation unit 217)
The vehicle body behavior calculation unit 217 acquires the acceleration of the vehicle body when the cargo handling vehicle 100 is traveling, using the acceleration sensor 105. The vehicle body behavior calculation unit 217 transmits the acquired acceleration to the work determination unit 230 and the control command generation unit 220.

(Control command generating unit 220)
The control command generating unit 220 receives information, for example, as described below, and controls the loading vehicle 100 in accordance with the plant model using the received information.

The control command generation unit 220 inputs the target speed set in the target node 302 and controls the loading vehicle 100 to run at this target speed. For example, when the loading vehicle 100 runs on a motor, the control command generation unit 220 generates a command value for the driving motor of the drive wheels of the loading vehicle 100, and feeds back the difference between the current command value and the response value of the driving motor, implementing control to reduce this difference.

The control command generating unit 220 also inputs the target travel direction set in the target link 303, and controls the loading vehicle 100 to travel along this target travel direction. For example, the control command generating unit 220 generates a command value for the steering motor of the loading vehicle 100, and feeds back the difference between the current command value and the response value of the steering motor, thereby implementing control to reduce this difference.

The control command generation unit 220 also inputs the task ID provided in the target node 302, and if the task identified by this task ID includes, for example, stopping the loading vehicle 100 at the target node 302, generates a brake command for the loading vehicle 100.

The control command generation unit 220 also inputs the task ID provided in the target node 302, and if the task identified by this task ID includes, for example, a loading operation of the loading vehicle 100 at the target node 302, generates a command to raise and lower the transfer device 102 and a command to move the loading member 103 forward and backward so that the tip of the loading member 103 can be inserted into the opening of the pallet 304 on which the target load is placed.

The control command generation unit 220 outputs the generated command values and commands to the drive units 221 to 224, which are described below.

(Steering motor drive unit 221)
The steering motor driving unit 221 drives a motor (steering motor) that transmits power to steered wheels of the cargo handling vehicle 100 in accordance with a command value input from the control command generating unit 220 .

(Travel motor drive unit 222)
The traveling motor driving unit 222 drives a motor (traveling motor) that transmits power to driving wheels equipped on the cargo handling vehicle 100 in accordance with a command value input from the control command generating unit 220 .

(Brake driving unit 223)
The brake driving unit 223 drives a motor that transmits power to a hydraulic pump that drives brake pads in accordance with a brake command input from the control command generating unit 220 in order to stop the rotation of the drive wheels of the cargo handling vehicle 100. The hydraulic pump is activated to press the brake pads against the drive wheels, thereby stopping the rotation of the drive wheels of the cargo handling vehicle 100.

(Transfer device/loading member motor drive unit 224)
The transfer device/loading member motor drive unit 224 drives a motor that transmits power to a hydraulic pump that drives the transfer device 102 and the loading member 103 equipped on the loading vehicle 100 in accordance with a command input from the control command generation unit 220. The transfer device 102 is raised and lowered by the operation of the hydraulic pump that drives the transfer device 102. The loading member 103 moves in the forward and backward directions by the operation of the hydraulic pump that drives the loading member 103.

(Overview of the work determination unit 230)
The work determination unit 230 determines the work to be performed by the cargo handling vehicle 100 at the target node 302. The process of the work determination unit 230 will be briefly described below. The process of the work determination unit 230 will be described in detail later.

The work determination unit 230 compares the work IDs of the current node 301 and the target node 302 determined by the nearest node calculation unit 215. The current node 301 and the target node 302 may be assigned multiple work IDs. In this case, the work determination unit 230 selects one work ID to compare depending on the state of the loading vehicle 100 (e.g., whether or not there is a load).

The work determination unit 230 determines whether the work ID in the target node 302 is different from the work ID in the current node 301. If the work ID in the target node 302 is different from the work ID in the current node 301, the work determination unit 230 requests the plant model switching unit 231 to switch the plant model for controlling the loading vehicle 100 to a plant model corresponding to the work specified by the work ID in the target node 302.

In addition, if the work ID in the target node 302 is different from the work ID in the current node 301 and is a work ID that has been preset in the route map management unit 205 of the traffic control unit 201, the work determination unit 230 can also request the plant model switching unit 231 to switch the plant model for controlling the loading vehicle 100 to a plant model corresponding to the work identified by the work ID in the target node 302. The work ID in the target node 302 that has been preset in the route map management unit 205 means that the work ID in the target node 302 is a work ID provided for the traveling node 302 in the route information stored by the route map management unit 205.

(Outline of the plant model switching unit 231)
The plant model switching unit 231 switches the plant model for controlling the loading vehicle 100. In response to a request from the work determination unit 230, the plant model switching unit 231 switches the plant model to a plant model corresponding to the work specified by the work ID in the target node 302, and outputs the switched plant model to the control command generation unit 220. Details of the processing by the plant model switching unit 231 will be described later.

(Overview of the plant model management unit 232)
Here, a brief explanation will be given of the plant model management unit 232 of the traffic control unit 201. The details of the processing of the plant model management unit 232 will be described later.

The plant model management unit 232 stores a plant model in advance for each task ID shown in FIG. 4. The plant model management unit 232 outputs a plant model in response to a request from the plant model switching unit 231 (i.e., a plant model corresponding to the task identified by the task ID in the target node 302) to the plant model switching unit 231.

The following describes in detail the processing of the work judgment unit 230 and plant model switching unit 231 of the on-board controller 107, and the plant model management unit 232 of the traffic control unit 201.

First, we will explain the details of the processing of the work judgment unit 230 and the plant model switching unit 231.

(Details of the work determination unit 230 and the plant model switching unit 231)
Fig. 5A is a flowchart showing the flow of processing by the work determination unit 230 and the plant model switching unit 231. The work determination unit 230 and the plant model switching unit 231 perform the processing shown in Fig. 5A after the loading vehicle 100 performs work at the current node 301 (closest node) and before performing work at the target node 302. The work determination unit 230 acquires the work (work ID at the target node 302) that the loading vehicle 100 performs at the target node 302, and determines whether or not it is necessary to switch the plant model for controlling the loading vehicle 100. The plant model switching unit 231 switches the plant model of the loading vehicle 100 in response to a request from the work determination unit 230.

In process 501, the work determination unit 230 obtains the current node 301 and the target node 302 of the loading vehicle 100 determined by the nearest node calculation unit 215 from the nearest node calculation unit 215.

In process 502, the work determination unit 230 acquires the state of the loading vehicle 100. The state of the loading vehicle 100 is, for example, whether the loading vehicle 100 is loaded (whether there is a load) and the posture of the loading vehicle 100. The work determination unit 230 acquires, for example, the mass of the current load on the loading vehicle 100 from the load weight calculation unit 216, and determines whether the loading vehicle 100 is loaded based on the acquired mass of the load. If the work determination unit 230 determines that the loading vehicle 100 is loaded, it acquires the mass of the load. Depending on the accuracy of the load sensor 106, for example, if the mass acquired from the load weight calculation unit 216 is less than 10 kg, the work determination unit 230 determines that the loading vehicle 100 is not loaded (no load).

In process 503, the work determination unit 230 compares the work IDs (FIG. 4) in the current node 301 and the target node 302. The work ID in the current node 301 is a work ID that identifies the work performed by the loading vehicle 100 in the current node 301. If multiple work IDs are assigned to the target node 302, the work determination unit 230 selects one work ID according to the state of the loading vehicle 100 acquired in process 502 (e.g., whether or not the loading vehicle 100 has a load) and sets the selected work ID as the work ID in the target node 302.

If the task ID in the current node 301 and the task ID in the target node 302 are different, proceed to process 504. If these task IDs are the same, end the process.

In process 504, the work determination unit 230 determines whether the distance Lv from the current position (self-position 311) of the loading vehicle 100 to the target node 302 is equal to or less than a predetermined distance L. The distance L is a threshold value that indicates whether the distance to the traveling node of the loading vehicle 100 is large, i.e., whether the loading vehicle 100 has reached the traveling node, and can be set arbitrarily in advance. When the distance Lv is equal to or less than the distance L, it indicates that the distance to the traveling node of the loading vehicle 100 is small and that the loading vehicle 100 has reached the traveling node. Note that the distance L may differ depending on the work ID and traveling node.

If the distance Lv is equal to or less than the distance L, the loading vehicle 100 has reached the target node 302, and the process proceeds to step 505. If the distance Lv is greater than the distance L, the loading vehicle 100 has not reached the target node 302, and the process ends.

In process 505, the work judgment unit 230 determines that since the loading vehicle 100 has reached the target node 302 (Lv≦L), it is necessary to switch the plant model for controlling the loading vehicle 100.

Figure 6 is a conceptual diagram showing loading vehicles 100a and 100b when switching plant models.

The following two cases will be described as examples of switching between plant models.
(i) When the loading vehicle 100a shifts from traveling forward in a straight line without a load (task ID 1-2 in FIG. 4) to traveling forward turning right without a load (task ID 3-2 in FIG. 4).
(b) When the loading vehicle 100b shifts from a right spin turn with a load (task ID 8-1 in FIG. 4) to straight backward travel with a load (task ID 2-1 in FIG. 4).

In FIG. 6, the loading vehicle 100a is in the situation (a) described above, and the task is to shift from forward straight-line travel without a load at the current node 301 (task ID 1-2 in FIG. 4) to forward right turn travel without a load at the target node 302 (task ID 3-2 in FIG. 4). For example, when the loading vehicle 100a is located within a circle of radius L centered on the target node 302, the task determination unit 230 determines that the distance Lv is less than or equal to distance L (Lv≦L) and the loading vehicle 100a has reached the target node 302, so that the plant model for controlling the loading vehicle 100a should be switched.

In FIG. 6, the loading vehicle 100b is in the situation (b) described above, and when it travels in a straight line (forward travel) and reaches the current node 601 from the travel node 600 (i.e., when the distance Lv from the self-position 311 of the loading vehicle 100b to the current node 601 becomes equal to or less than the distance L), it loads the pallet 304 as cargo at the current node 601 and executes a right spin turn with cargo (task ID 8-1 in FIG. 4) at the current node 601. Then, when the loading vehicle 100b completes the right spin turn with cargo, it executes straight backward travel with cargo (task ID 2-1 in FIG. 4). The target link 603 of the loading vehicle 100b is a link that connects the current node 601 and the target node 602.

In process 505 of FIG. 5, the work determination unit 230 checks whether the self attitude 312 has become the target attitude at the current node 601 when the loading vehicle 100b performs a spin turn at the current node 601, and may determine to switch the plant model for controlling the loading vehicle 100b when the self attitude 312 of the loading vehicle 100b has become the target attitude at the current node 601. In this case, the work determination unit 230 determines to switch the plant model that performs the spin turn to a plant model that travels straight ahead. For example, when the plant model is switched, the loading vehicle 100b performs the work of traveling backward in a straight line with a load (work ID 2-1 in FIG. 4) at the current node 601.

For example, when the loading vehicle 100b executes a right spin turn, if the self attitude 312 with respect to the target link 603 (angle φv of the loading vehicle 100b in the forward/rearward direction _YV with respect to the target link 603) is equal to or less than the angle φ, the work determination unit 230 determines that the self attitude 312 has become the target attitude at the current node 601, and determines to switch the plant model for controlling the loading vehicle 100b. The angle φ is a threshold value that indicates whether or not the difference between the self attitude 312 of the loading vehicle 100b and the target attitude is large, and can be determined arbitrarily in advance.

If the self-attitude 312 (angle φv) is equal to or less than the angle φ, the work determination unit 230 switches the plant model and causes the loading vehicle 100b to perform straight-line backward travel with a load at the current node 601. In this way, the work determination unit 230 can switch the plant model that performs a spin turn to a plant model that performs straight-line backward travel according to the self-attitude 312 of the loading vehicle 100b, and transition the work of the loading vehicle 100b from a spin turn to straight-line backward travel.

In process 505 of FIG. 5A, when the work ID in the target node 302 is different from the work ID in the current node 301 (process 503) and the loading vehicle 100 has reached the target node 302 (Lv≦L), the work determination unit 230 determines to switch the plant model for controlling the loading vehicle 100 and requests the plant model switching unit 231 to switch the plant model. The work determination unit 230 transmits a request to the plant model switching unit 231 to switch the plant model for controlling the loading vehicle 100 to a plant model corresponding to the work identified by the work ID in the target node 302.

The work determination unit 230 may also check whether the work ID in the target node 302 is a work ID that has been preset in the route map management unit 205, and if the work ID in the target node 302 is different from the work ID in the current node 301 and is a work ID that has been preset in the route map management unit 205, request the plant model switching unit 231 to switch the plant model.

Figure 5B is a flowchart showing another process flow of the work determination unit 230 and the plant model switching unit 231. In the flowchart shown in Figure 5B, process 505a is executed instead of process 505 in the flowchart shown in Figure 5A. Only process 505a will be described for the flowchart shown in Figure 5B.

In process 505a, the work determination unit 230 checks whether the work ID in the target node 302 is a work ID that has been preset in the route map management unit 205, and if the work ID in the target node 302 is a work ID that has been preset in the route map management unit 205, it requests the plant model switching unit 231 to switch the plant model.

If the task ID in the target node 302 is different from the task ID in the current node 301, but is not a task ID preset in the route map management unit 205, the task identified by this task ID in the target node 302 is not a task to be performed by the loading vehicle 100. It can be assumed that this task ID in the target node 302 is a task ID obtained by mistake. In this case, the loading vehicle 100 stops the task. In this case, the control command generation unit 220 performs control to stop the loading vehicle 100, for example, by outputting a command to the travel motor drive unit 222 to stop the loading vehicle 100. In this way, the loading vehicle system 200 according to this embodiment can confirm that the task identified by the task ID in the target node 302 is definitely a task to be performed by the loading vehicle 100.

In process 506 of FIG. 5A, the plant model switching unit 231 switches the plant model to a plant model corresponding to the work specified by the work ID in the target node 302 in response to a request from the work determination unit 230. When the plant model switching unit 231 receives a request to switch the plant model from the work determination unit 230, it acquires the plant model corresponding to the work specified by the work ID in the target node 302 from the plant model management unit 232 of the traffic control unit 201, and outputs the acquired plant model to the control command generation unit 220. In this way, the plant model switching unit 231 switches the plant model for controlling the loading vehicle 100.

In this embodiment, the plant model switching unit 231 is provided in the vehicle controller 107, but may be provided in the traffic control unit 201. When the traffic control unit 201 is provided with the plant model switching unit 231, the vehicle controller 107 requests the plant model switching unit 231 of the traffic control unit 201 to switch the plant model via the communication device 202 and the communication device 203 of the traffic control unit 201.

The plant model in this embodiment will be explained together with the explanation of the plant model management unit 232.

Next, we will explain the plant model management unit 232 of the traffic control unit 201.

(Details of the plant model management unit 232)
The plant model management unit 232 stores a plant model for each task ID. The plant model is a model that indicates the operating characteristics when the autonomously controlled cargo handling vehicle 100 performs a task. In other words, the plant model corresponding to a task specified by a task ID indicates the operating characteristics of the cargo handling vehicle 100 when performing the task specified by the task ID. The plant model is expressed, for example, by a transfer function that indicates the relationship between a command value of the steering motor and its response value, and an equation of motion for the cargo handling vehicle 100. The plant model management unit 232 manages the transfer function and the model (mathematical formula) of the equation of motion and parameters (coefficient values, etc.) for each task ID.

In this embodiment, the plant model management unit 232 is provided in the traffic control unit 201, but may also be provided in the on-board controller 107.

(Plant model expressed as a transfer function)
In the following, the transfer functions of the travel motor and steering motor of the cargo handling vehicle 100 will be described as an example of the plant model.

Fig. 7A is a diagram showing a transfer function G ₁ (s) of the traveling motor. Fig. 7B is a diagram showing an example of the relationship between a command value 701 and a response value 702 of the traveling motor. The transfer function G ₁ (s) of the traveling motor shows the relationship between a command value 701 (vr) and a response value 702 (vy) of the traveling motor.

The plant model management unit 232 uses the traveling data acquired in advance to construct the transfer function G ₁ (s) shown in Fig. 7A from the relationship between the command value 701 and the response value 702 of the traveling motor as shown in Fig. 7B. For example, it is assumed that the relationship between the command value 701 (vr) of the traveling motor and the response value 702 (vy) can be expressed using the transfer function G ₁ (s) shown in equation (1-1).

In equation (1-1), s is the Laplace transform variable, and a, b, and c are arbitrary constants.

Figure 8A is a diagram showing the transfer function of the steering motor. Figure 8B is a diagram showing an example of the relationship between the command value 801 and the response value 802 of the steering motor. Figure 8B also shows the relationship between the command value 701 and the response value 702 of the travel motor shown in Figure 7B. The transfer function of the steering motor shows the relationship between the command value 801 (θr) and the response value 802 (θy) of the steering motor.

The plant model management unit 232 uses the driving data acquired in advance to construct the transfer function shown in FIG. 8A from the relationship between the steering motor command value 801 and the response value 802 as shown in FIG. 8B.

In the case where both the steering motor and the traveling motor are configured with hydraulic mechanisms (e.g., hydraulic motors), the hydraulic oil used to drive the traveling motor may be distributed to the steering motor when the steering motor is driven during a turning operation. In this embodiment, it is assumed that the steering motor response value 802 (θy) is affected by the steering motor command value 801 (θr) and the traveling motor response value 702 (vy), and the plant model management unit 232 creates a two-input, one-output transfer function. In this embodiment, it is assumed that the transfer function G ₂ (s) representing the relationship between the steering motor command value 801 (θr) and the steering motor response value 802 (θy) can be expressed by the linear transfer function of the formula (1-2), and the transfer function G ₃ (s) representing the relationship between the traveling motor response value 702 (vy) and the steering motor corrected command value θu can be expressed by the linear transfer function of the formula (1-3).

In equations (1-2) and (1-3), K1 and K2 are the gains (steady-state value/command value), and t1 and t2 are the time constants (the difference between the time when the response value reaches 63.2% (=1-1/e) of the steady-state value and the time when the command begins).

In the cargo handling vehicle system 200 according to this embodiment, for example, by using different transfer functions (plant models) when the cargo handling vehicle 100 travels in a straight line and when it turns, it is possible to realize control with high tracking accuracy for the target route. For example, it is possible to use different transfer functions according to the following conditions.

Condition 1) Transfer function G ₁ (s)
When the turning radius R at the target node 302 is greater than radius R0, it is regarded as straight line driving, and only the transfer function G ₁ (s) indicating the relationship between the command value 701 (vr) and the response value 702 (vy) of the driving motor is used.

Condition 2) Transfer function _G2 (s)+ _G3 (s)
When the turning radius R at the target node 302 is less than radius R0, the vehicle is considered to be turning, and a transfer function _G2 (s)+ _G3 (s) is used, which represents the relationship between the steering motor command value 801 (θr) and the response value 802 (θy) taking into account the steering motor corrected command value (θu).

(Plant model expressed by the equation of motion)
In the following, the equation of motion of the cargo handling vehicle 100 will be described as an example of the plant model.

When controlling the speed and angular velocity of the drive wheels of the loading vehicle 100, it is necessary to define the equation of motion of the vehicle, such as the vehicle dynamics of the loading vehicle 100. The equation of motion of the vehicle is determined using geometric conditions between the wheels, such as the wheelbase and tread. If the loading vehicle 100 is moving forward when the loading member 103 is located forward in the direction of travel, then the loading vehicle 100 is rear-wheel driven when moving forward, and front-wheel driven when moving backward (when the loading member 103 is located backward in the direction of travel). For this reason, even when traveling in a straight line, for example, different equations of motion must be prepared for forward and backward travel. Below, examples of equations of motion (plant models) are explained for the loading vehicle 100 moving forward (rear-wheel drive) and backward (front-wheel drive).

Figure 9 is a diagram showing a cargo handling vehicle 100 traveling forward (rear-wheel drive). Figure 10 is a diagram showing a cargo handling vehicle 100 traveling backward (front-wheel drive). In Figures 9 and 10, X and Y indicate the self-position 311 of the cargo handling vehicle 100. That is, X indicates the X-direction position of the cargo handling vehicle 100 in the reference coordinate system 310, and Y indicates the Y-direction position of the cargo handling vehicle 100 in the reference coordinate system 310. In addition, in Figures 9 and 10, l indicates the wheelbase, tr indicates the tread, θ indicates the rotation angle of the vehicle body, δ indicates the rotation angle of the drive wheels, Vd indicates the forward speed of the drive wheels, and V indicates the forward speed of the cargo handling vehicle 100. Although not shown, if the radius of the drive wheels is rd and the rotation speed of the drive wheels is ω, the forward speed Vd of the drive wheels can be expressed as Vd = ω x rd.

Generally, for vehicles with a maximum speed of about 10 km/h, such as the cargo handling vehicle 100, a kinematic model (a model derived from geometric relationships) that does not take into account the wheel slip angle can provide sufficiently accurate control. For this reason, in this embodiment, in order to reduce the number of control elements, the plant model is defined using the kinematic model.

In this embodiment, the equation of motion for the cargo handling vehicle 100 is defined assuming that the slip angles and speeds of the four wheels of the cargo handling vehicle 100 are equal to each other.

In the case of forward travel (rear-wheel drive) shown in Figure 9, the equations of motion for the loading vehicle 100 can be expressed by the equations of motion for rear-wheel drive, equations (1-4), (1-5), and (1-6), shown below.

In the case of reverse driving (front-wheel drive) shown in Figure 10, the equations of motion of the loading vehicle 100 can be expressed by the equations of motion for front-wheel drive, equations (1-7), (1-8), and (1-9), shown below.

In the cargo handling vehicle system 200 according to this embodiment, for example, by using different equations of motion (plant models) when the cargo handling vehicle 100 travels forward and when it travels backward, it is possible to achieve control with high tracking accuracy for the target route. For example, it is possible to use different equations of motion according to the following conditions.

Condition 1) Equation of motion for rear-wheel drive When the loading vehicle 100b in FIG. 6 travels from the traveling node 600 to the current node 601 (i.e., when the distance Lv from the self-position 311 of the loading vehicle 100b to the current node 601 is greater than the distance L), the loading vehicle 100b travels forward, so the equation of motion for rear-wheel drive is used.

Condition 2) Equation of motion for front-wheel drive When the loading vehicle 100b is performing a right spin turn at the current node 601 in FIG. 6, if its own attitude 312 (angle φv) with respect to the target link 603 is equal to or less than the angle φ, the loading vehicle 100b will end the right spin turn and begin traveling backwards, so the equation of motion for front-wheel drive is used.

In the loading vehicle system 200 according to the present embodiment, the loading vehicle 100 automatically acquires the work to be performed and automatically switches the plant model according to the acquired work (work ID), so that the traveling of the loading vehicle 100 can be controlled with high precision. In addition, when the loading vehicle system 200 according to the present embodiment is used, the traveling of the loading vehicle 100 can be controlled with high precision, so that, for example, the deviation from the desired route during the traveling of the loading vehicle 100 can be reduced, and the route when moving to the target position can be shortened, so that the transportation efficiency of the loading vehicle 100 can be improved. Furthermore, when the loading vehicle system 200 according to the present embodiment is used, the arrangement of the cargo in the warehouse can be designed according to the plant model, so that there are fewer constraints on the arrangement of the cargo in the warehouse, and not only the transportation efficiency of the loading vehicle 100 can be improved, but also the arrangement of the cargo in the warehouse can be optimized.

A loading vehicle system 200 according to a second embodiment of the present invention will be described with reference to Figures 11 to 12B. The loading vehicle system 200 according to this embodiment has almost the same configuration as the loading vehicle system 200 according to the first embodiment, but differs from the loading vehicle system 200 according to the first embodiment in that the traffic control unit 201 includes a learning processing unit. The learning processing unit includes a learning data selection unit 1100, a learning unit 1101, and a database update unit 1102, and updates the plant model corresponding to the work identified by the work ID.

The following describes the learning processing units provided in the loading vehicle system 200 according to this embodiment: the learning data selection unit 1100, the learning unit 1101, and the database update unit 1102.

Figure 11 is a functional block diagram showing the configuration of a cargo handling vehicle system 200 according to this embodiment. The cargo handling vehicle system 200 according to this embodiment is configured such that the traffic control unit 201 further includes a learning processing unit (learning data selection unit 1100, learning unit 1101, and database update unit 1102) in the cargo handling vehicle system 200 according to embodiment 1 (Figure 2).

(Learning Data Selection Unit 1100)
The learning data selection unit 1100 inputs the task ID of the current node 301 of the loading vehicle 100 from the task determination unit 230 of the on-board controller 107. Furthermore, the learning data selection unit 1100 inputs command values (e.g., command values to the steering motor drive unit 221, the travel motor drive unit 222, the brake drive unit 223, and the transfer device/loading member motor drive unit 224) for controlling the loading vehicle 100 at the current node 301 and the response values of the loading vehicle 100 to these command values (response values of each drive unit to the command value) from the control command generation unit 220. The learning data selection unit 1100 outputs the input task ID, command value, and response value of the current node 301 to the learning unit 1101 as learning data.

(Learning unit 1101)
The learning unit 1101 inputs the work ID at the current node 301, the command value to the loading vehicle 100, and the response value of the loading vehicle 100 to the command value from the learning data selection unit 1100, and updates the plant model by machine learning using these command values and response values as learning data.

Below, as an example, a description is given of a case in which the learning unit 1101 uses a command value for the steering motor of the loading vehicle 100 and a response value from the steering motor to update the parameters of a plant model (a plant model represented by a transfer function) through machine learning using feedforward control.

The learning unit 1101 acquires from the learning data selection unit 1100 the command value θr and the response value θy of the steering motor when the loading vehicle 100 turns while traveling forward or when the loading vehicle 100 turns while traveling backward. In this embodiment, the learning unit 1101 obtains a transfer function that indicates the responsiveness of the loading vehicle 100, which is different for turning while traveling forward and turning while traveling backward. The learning unit 1101 uses the obtained transfer function to obtain an inverse response, thereby reducing the error between the command value and the response value and improving the accuracy of the path following function.

Figure 12A is a diagram showing a transfer function that indicates the responsiveness of the steering motor. The relationship between the steering motor command value θr and the response value θy is expressed by the transfer function P(s). It is assumed that the transfer function P(s) is expressed by the formula (2-1).

In equation (2-1), s is the Laplace transform variable, and a, b, and c are arbitrary constants (parameters).

In the learning unit 1101, the relationship between the steering motor command value θr and the response value θy is expressed by the transfer function P(s) in FIG. 12A, and the transfer function P(s) is expressed by the formula (2-1).

The learning unit 1101 can obtain the inverse response P ⁻¹ (s) using the transfer function P(s) shown in equation (2-1), and can configure the feedforward control F(s) shown in FIG. 12B.

Figure 12B shows feedforward control using a transfer function indicating the responsiveness of the steering motor obtained by the learning unit 1101.

It is desirable for the steering motor response value θy to be as responsive as possible to the command value θr. Therefore, consider a feedforward control F(s) as shown in FIG. 12B. The output of the feedforward control F(s) is the corrected command value θu of the steering motor. For this reason, the corrected command value θu of the steering motor calculated by the control command generation unit 220 is sent to the steering motor drive unit 221 (FIG. 11).

The feedforward control F(s) is given, for example, by equation (2-2).

In equation (2-2), ωc and ηc are arbitrary constants.

Equation (2-2) can be expressed as equation (2-4) using the transfer function L(s) shown in equation (2-3).

As shown in FIG. 12B, the response value θy of the steering motor relative to the command value θr can be arbitrarily designed using the transfer function L(s) shown in equation (2-3). Therefore, if the transfer function P(s) showing the relationship between the steering motor command value θr and the response value θy is determined in advance, the steering motor can be adjusted to provide the desired response.

The learning unit 1101 acquires a data set of the command value θr and response value θy of the steering motor as learning data, and by using the command value θr and the response value θy, it is possible to appropriately calculate the transfer function P(s) by the least squares method or system identification focusing on frequency response, and to successively update the transfer function P(s). The learning unit 1101 can update the parameters of the plant model (the plant model represented by the transfer function) by calculating and updating the three parameters a, b, and c of the transfer function P(s) shown in equation (2-1).

(Database update unit 1102)
The database update unit 1102 acquires the parameters of the plant model updated by the learning unit 1101 (for example, parameters a, b, and c of the transfer function P(s) shown in equation (2-1)) from the learning unit 1101, and updates the values of the parameters of the plant model for each task ID. The database update unit 1102 outputs the updated plant model to the plant model management unit 232.

The plant model management unit 232 stores the plant model updated by the database update unit 1102 for each work ID.

The loading vehicle system 200 according to this embodiment is equipped with a learning processing unit (learning data selection unit 1100, learning unit 1101, and database update unit 1102) that updates the plant model, and can automatically generate appropriate learning data for each job ID to improve the accuracy of the plant model. Therefore, the loading vehicle system 200 according to this embodiment can be equipped with a tuning-less control system, which can reduce engineering costs when introducing the system.

The present invention is not limited to the above-mentioned examples, and various modifications are possible. For example, the above-mentioned examples have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to an embodiment having all of the configurations described. It is also possible to replace part of the configuration of one example with the configuration of another example. It is also possible to add the configuration of another example to the configuration of one example. It is also possible to delete part of the configuration of each example, or to add or replace other configurations.

100, 100a, 100b...loading vehicle, 101...vehicle frame, 102...transfer device, 103...loading member, 104...external sensor, 105...acceleration sensor, 106...load sensor, 107...vehicle controller, 200...loading vehicle system, 201...traffic control unit, 202...vehicle controller communication device, 203...traffic control unit communication device, 204...operation management unit, 205...route map management unit, 206...traveling path generation unit, 214...self-position and attitude calculation unit, 215...nearest node calculation unit, 216...load load calculation unit, 217...vehicle body behavior calculation unit, 220...control command generation unit, 221...steering motor drive unit, 222...traveling motor drive unit, 223...brake drive unit, 224...transfer device/loading member Motor drive unit, 230...work judgment unit, 231...plant model switching unit, 232...plant model management unit, 301...traveling node (current node), 302...traveling node (target node), 303...link (target link), 304...pallet, 305...segment, 306...terminal node, 307...shelf, 310...reference coordinate system, 311...self-position, 312...self-attitude, 600...traveling node, 601...current node, 602...target node, 603...target link, 701...traveling motor command value, 702...traveling motor response value, 801...steering motor command value, 802...steering motor response value, 1100...learning data selection unit, 1101...learning unit, 1102...database update unit.

Claims (5)

The loading and unloading system includes a loading member on which a load to be transported by the loading and unloading vehicle is placed, a control command generating unit, a travel path generating unit, a nearest node calculating unit, an operation determining unit, and a plant model switching unit,
the control command generating unit controls the cargo handling vehicle in accordance with a plant model that indicates an operating characteristic of the cargo handling vehicle when performing work;
the travel route generation unit creates a travel route for the cargo handling vehicle, the travel route being composed of a plurality of nodes indicating positions through which the cargo handling vehicle passes and one or more links connecting two of the nodes, and each of the nodes being provided with a work ID indicating a work to be performed by the cargo handling vehicle;
the nearest node calculation unit inputs the position of the cargo handling vehicle and the travel route, determines, among the nodes constituting the travel route, a node located closest to the position of the cargo handling vehicle as a current node, and determines, among the nodes constituting the travel route, a destination node which is a node that the cargo handling vehicle should pass through next after the current node;
The task determination unit compares the task ID at the current node with the task ID at the goal node;
a plant model switching unit, when the work ID at the current node and the work ID at the target node are different from each other, switches the plant model to the plant model corresponding to the work specified by the work ID at the target node.
A loading vehicle system comprising: The operation determination unit determines whether or not a distance from a position of the cargo handling vehicle to the target node is equal to or shorter than a predetermined distance;
a plant model switching unit, when the work ID at the current node and the work ID at the target node are different from each other and the distance from the position of the loading vehicle to the target node is equal to or shorter than the predetermined distance, switches the plant model to the plant model corresponding to the work specified by the work ID at the target node.
2. The loading vehicle system of claim 1. a route map management unit that stores data of a two-dimensional map of a range in which the cargo handling vehicle travels and in which the work ID is preset;
The plant model switching unit is
when the work ID at the current node and the work ID at the target node are different from each other and the work ID at the target node is the work ID set in the route map management unit, switch the plant model to the plant model corresponding to the work specified by the work ID at the target node;
3. A loading vehicle system according to claim 1 or 2. a control command generating unit, when the work ID at the current node and the work ID at the target node are different from each other and the work ID at the target node is not the work ID set in the route map managing unit, performs control to stop the loading vehicle.
4. The loading vehicle system according to claim 3. a learning processing unit that updates the plant model,
the learning processing unit inputs a command value for controlling the cargo handling vehicle and a response value of the cargo handling vehicle to the command value from the control command generation unit, and updates parameters of the plant model by machine learning using the command value and the response value.
2. The loading vehicle system of claim 1.

Images