JP2022159718A - Control device and control system - Google Patents

Abstract

To realize a safe traffic flow by suppressing interference between moving objects with each other in an interference area even if a state regarding a movement of a moving object is different from the predicted state.SOLUTION: A control device 10 receives moving object information representing a state of each of vehicles 7 including an autonomous vehicle; detects whether or not an abnormality related to a movement of a manual vehicle has occurred from a change in a state of the manual vehicle that may interfere with the autonomous vehicle at an interference point X among each of the vehicles 7 which has received the moving object information; and generates a reinforced virtual traffic rule for the autonomous vehicle that is more restricted than a virtual traffic rule generated before the abnormality was detected so that the manual vehicle is less likely to interfere with the autonomous vehicle at the interference point X when the abnormality is detected.SELECTED DRAWING: Figure 23

Description

The present invention relates to control devices and control systems.

In Patent Document 1, an integrated control server that successively and integrally manages arbitrary statuses of a plurality of unmanned vehicles traveling along a prescribed travel route through communication controls two or more unmanned vehicles at intersections where the travel routes intersect each other. When the vehicles are approaching, it is determined whether the vehicles will enter at the timing of merging and interference, and if it is determined that the vehicle will enter the intersection, On the other hand, a traffic control method is disclosed that determines passage priority based on a predetermined order.

Patent Document 2 discloses a vehicle management system in which an unmanned vehicle, a manned vehicle, and a control station that controls and controls them are connected so as to be able to communicate with each other. When an abnormality in the position notification is determined, a command is generated to output a warning requesting the driver of the manned vehicle to stop. It is disclosed to generate an abnormal unmanned vehicle no-entry area around the vehicle.

JP 2019-46013 A JP 2020-197885 A

The vehicle travel control method in Patent Literature 1 is a technique based on the premise that the vehicle operates within the prediction range of the integrated control server. However, when the vehicle is actually running, for example, the vehicle may suddenly decelerate due to detection of an obstacle. Therefore, in the vehicle travel control method disclosed in Patent Literature 1, if the vehicle behaves in a way that is not predicted by the integrated control server, the vehicles may interfere with each other at the intersection.

In addition, in the vehicle management system in Patent Document 2, when the control station recognizes an abnormality in the position notification from the manned vehicle, it is an area centered on manned vehicles, and an abnormality that prohibits unmanned vehicles from entering the area When unmanned vehicle entry prohibited area is generated. Therefore, for example, when a manned vehicle is approaching an unmanned vehicle that has already entered the interference area, no matter how much an unmanned vehicle entry prohibition area is generated for the manned vehicle, the manned vehicle cannot enter the interference area. Vehicles and unmanned vehicles may interfere.

Furthermore, since an unmanned vehicle cannot enter an unmanned vehicle-prohibited area during an emergency, for example, if an unmanned vehicle-prohibited area is generated for a manned vehicle that is about to pass through an intersection, the unmanned vehicle must stop at a position away from the stop line drawn in front of the intersection, which hinders traffic flow.

On the other hand, even after the manned vehicle has passed through the intersection, the intersection is included in the unmanned vehicle entry prohibition area during abnormal conditions for a while, so the unmanned vehicle cannot start immediately. In other words, the unmanned vehicle continues to stop before the intersection, which also hinders the traffic flow.

The present invention has been made in view of the above-described problems, and is a safety device that suppresses interference between moving bodies in an interference area even when the state of movement of the moving bodies is different from the predicted state. An object of the present invention is to provide a control device and a control system that can realize a smooth traffic flow.

2. The control device according to claim 1, comprising: a receiving unit for receiving moving object information representing the state of each moving object including an autonomous moving object that receives constraints defining movement time and space; Abnormal movement of the approaching moving body occurs due to a change in the state of the approaching moving body, which has the possibility of approaching the autonomous moving body within a predetermined range in the interference area, among the moving bodies that have received the body information. and when an abnormality is detected by the abnormality detection unit, the approaching moving object and the and a generation unit that generates, for the autonomous mobile body, an enhanced constraint that is the constraint that strengthens the constraint so as to make it difficult for interference to occur in the interference area.

In the invention according to claim 2, the abnormality detection unit determines whether the state value regarding the state of the approaching moving object is within a prediction range set using the moving body information of the approaching moving object received by the receiving unit. If not, it is detected that an abnormality related to the movement of the approaching moving object has occurred.

In the invention according to claim 3, the anomaly detection unit detects the speed of the approaching moving body, the acceleration of the approaching moving body, the passage time for the approaching moving body to pass through the interference area, and the position of the approaching moving body. When the state value is not included in the predicted range set for the state of at least one of the approaching moving bodies, it is detected that an abnormality related to the movement of the approaching moving body has occurred.

In the invention according to claim 4, the abnormality detection unit detects the speed of the approaching moving body, the acceleration of the approaching moving body, the passage time for the approaching moving body to pass through the interference area, and the position of the approaching moving body. If the measurement accuracy of the state values related to the state of at least one of the approaching moving bodies is lower than a predetermined reference accuracy, even if the state values are within the prediction range, Detect that an abnormality has occurred.

In the invention according to claim 5, the generating unit sets the entry prohibition time for prohibiting entry into the interference area to be longer than the constraint condition generated before the abnormality is detected by the abnormality detection unit. The set strengthening constraint condition is generated for the autonomous mobile body.

In the invention according to claim 6, the generating unit may generate the above-mentioned in the strengthened constraint extended with respect to the prohibition time for entering the interference area in the constraint generated before the abnormality is detected by the abnormality detection unit. The extension time of the entry prohibition time to the interference area is adjusted using the shape of the interference area.

In the invention according to claim 7, when the approaching moving body is a connected moving body that receives the constraint generated by the generating unit and moves according to the received constraint, the generating unit generates for the autonomous mobile body and the connected mobile body, respectively, the strengthened constraint set so that the entry prohibition time to the interference area overlaps for each of the autonomous mobile body and the connected mobile body.

In the eighth aspect of the invention, as the number of detections of anomalies related to the movement of the approaching moving body increases, the generation unit generates more than the constraint conditions generated before the anomaly was detected by the anomaly detection unit. The strong constraint condition is generated for the autonomous mobile body, in which a long entry prohibition time into the interference area is set.

In the invention according to claim 9, the generation unit is set so that the range in which the autonomous mobile body is prohibited from entering is wider than the constraint conditions generated before the abnormality is detected by the abnormality detection unit. The strengthening constraint is generated for the autonomous mobile body.

In the invention according to claim 10, as the number of detections of anomalies relating to the movement of the approaching moving body increases, the generation unit generates more constraints than the constraint conditions generated before the anomaly was detected by the anomaly detection unit. The strengthened constraint is generated for the autonomous mobile body by setting a wide range of the interference area that prohibits the autonomous mobile body from entering.

A control system according to claim 11, in an interference area, includes moving object information representing a state of an approaching moving object that may approach within a predetermined range from an autonomous moving object that receives constraints defining travel time and space. , and an acquisition device for acquiring mobile body information representing the state of the autonomous mobile body, a receiving unit for receiving the mobile body information of the autonomous mobile body and the approaching mobile body acquired by the acquisition device from the acquisition device, an anomaly detection unit for detecting whether an anomaly related to the movement of the approaching moving object has occurred from a change in the state of the approaching moving object represented by the moving object information received by the receiving unit; is detected, the constraint is stronger than the constraint generated before the abnormality is detected by the abnormality detection unit so that the approaching moving body is less likely to interfere with the interference area. and a generation unit that generates a strengthening constraint condition for the autonomous mobile body.

According to the present invention, it is possible to realize a safe traffic flow that suppresses interference between mobile bodies in the interference area even when the state of movement of the mobile bodies is different from the predicted state.

It is a figure which shows the system configuration example of a control system. It is a figure which shows the functional structural example in the control apparatus of an observance vehicle. FIG. 3 is a diagram showing a functional configuration example of a control device for a semi-compliant vehicle; It is a figure which shows the functional structural example in the control apparatus of a non-compliance vehicle. FIG. 5 is a diagram showing an example of differences in driving functions for each vehicle type; It is a figure which shows the functional structural example of a collection apparatus and a control apparatus. It is a figure which shows the example of the content of a control driving|running route map. It is a figure which shows an example of an interference point. It is a figure which shows an example of interference point information. FIG. 4 is a diagram showing an example of a global route; It is a figure which shows an example of a driving route. It is a figure which shows the principal part structural example of the electric system in a control apparatus. FIG. 2 is a diagram showing a configuration example of a main part of an electrical system in a vehicle control device; 4 is a flowchart showing an example of the flow of control processing executed by a control device; FIG. 4 is a diagram showing an example of vehicle interference time at an interference point; FIG. 4 is a schematic diagram that schematically illustrates a virtual traffic rule at an interference point in which state s _x =0; FIG. 4 is a schematic diagram that schematically illustrates a virtual traffic rule at an interference point in which state s _x =1; FIG. 4 is a schematic diagram illustrating cooperation of interference points; 4 is a flowchart showing an example of the flow of virtual traffic rule generation processing; FIG. 2 illustrates an example of traffic flow caused by vehicle speed changes; FIG. 10 is a diagram showing an example of a virtual traffic rule for automatic vehicles with respect to changes in speed of manual vehicles; FIG. 10 is a diagram showing a state example in which the speed of the vehicle is not within the prediction range; FIG. 4 is a diagram showing an example of enhanced virtual traffic rules; FIG. 4 is a diagram showing a state example in which the speed of a vehicle is within a prediction range; It is a figure which shows an example of the prohibition area. FIG. 10 is a diagram showing an example of enhanced virtual traffic rules in a situation where passage priority is set to priority; FIG. 10 is a diagram showing an example of constraint conditions when vehicles approach each other in a situation in which passage priority is set to priority. FIG. 10 is a diagram showing an example of enhanced virtual traffic rules in a situation where passage priority is set to non-priority; FIG. 10 is a diagram showing an example of constraint conditions when vehicles approach each other in a situation in which passage priority is set to non-priority;

Hereinafter, embodiments will be described with reference to the drawings. The same constituent elements and the same processing are given the same reference numerals throughout the drawings, and overlapping descriptions are omitted.

FIG. 1 is a diagram showing a system configuration example of a control system 1 according to this embodiment. The control system 1 includes a vehicle 7 to be controlled by the control system 1, a plurality of wireless communication devices 3 and collection devices 9 respectively installed along the route 6, and a control device 10, each wireless communication device 3 and The collection device 9 is connected to the control device 10 through the communication network 5 .

Vehicles 7 are classified into a plurality of types according to differences in driving functions, which are functions related to driving provided in vehicles 7 . For example, the vehicle 7 is equipped with wireless equipment, and by wirelessly connecting to any one of the wireless communication devices 3 installed at the nearest location during travel, data communication is performed with the control device 10 through the communication network 5. Vehicles 7 are classified as connected vehicles, and vehicles 7 that do not have wireless equipment and cannot perform data communication with the control device 10 are classified as non-connected vehicles.

The wireless communication device 3 serves as a data relay device between the connected vehicle and the control device 10 . The wireless communication device 3 is not restricted in its installation location as long as it is within the range where wireless communication with the connected vehicle is possible. It is preferable to install it as close as possible to the roadway. In addition, there is no restriction on the number of installed wireless communication devices 3 .

The communication network 5 transmits various information of the vehicle 7 collected by the wireless communication device 3 and the collection device 9 to the control device 10, and the control device 10 designates the control information of the connected vehicle generated by the control device 10. It transmits to the wireless communication device 3 . The control information may be transmitted to all wireless communication devices 3 according to an instruction from the control device 10 and notified to each connected vehicle. Note that the communication network 5 may be a wired line or a wireless line.

The control device 10 does not reduce the traffic efficiency between the vehicles 7, and the vehicle 7 approaches a predetermined range (hereinafter referred to as "interference range") where the possibility of contact between the vehicles 7 is recognized. It is a device that performs traffic control to avoid interference at an interference point X. Here, "interference" means not only a situation in which the vehicles 7 come into contact with each other, but also a situation in which the vehicles 7 come close to each other within the range of interference.

The control device 10 receives mobile body information representing the state of the vehicle 7 from each vehicle 7 to be controlled, and sets constraint conditions (hereinafter referred to as "virtual traffic rules") that restrict the determination of movement performed by each vehicle 7. ) is generated for each vehicle 7 having a function of receiving constraints and is transmitted to each vehicle 7 having a function of receiving constraints. As a result, traffic control is performed so that the vehicles 7 do not interfere with each other at the interfering point X.

Here, the "state of the vehicle 7" refers to the movement of the vehicle 7, the details of the control performed on the vehicle 7, and the operational status of each function of the vehicle 7, which can be measured and collected. It is a collection of information that represents items related to each item.

Note that the virtual traffic rule transmitted from the control device 10 to the vehicle 7 is an example of a constraint condition for the vehicle 7, and the details will be described later.

The control device 10 can receive mobile object information through the wireless communication device 3 in the case of connected vehicles. unable to receive.

Therefore, the control device 10 acquires the moving body information of the non-connected vehicle through the collection device 9 .

The collection device 9 uses, for example, an imaging unit that optically captures images, a UWB (Ultra Wide Band) radar, or the like to capture or measure the traffic environment around the non-connected vehicle and the non-connected vehicle in the lane 8, and the non-connected vehicle It is a device that collects mobile information related to

Connected vehicles are separated into three types, compliant, semi-compliant, and non-compliant, according to their different driving functions.

A compliant vehicle is one that receives virtual traffic rules from the control device 10, meets the received virtual traffic rules while referring to traffic environment information obtained from sensors or the like attached to the vehicle 7, and interferes with the other vehicles 7. It is a vehicle 7 equipped with an automatic driving function that runs while controlling the vehicle 7 based on its own judgment so as to autonomously avoid Note that automatic driving according to the present embodiment refers to driving classified into a category of level 3 or higher in which the automatic driving function, not the driver, is responsible for driving. A compliant vehicle is an example of an autonomous mobile body that moves according to the judgment of the vehicle 7 itself.

A quasi-compliant vehicle receives the virtual traffic rules from the control device 10, but unlike the compliant vehicle, the vehicle 7 itself does not control traveling so as to satisfy the virtual traffic rules, but receives the virtual traffic rules. It is a vehicle 7 driven by a driver who satisfies the virtual traffic rules and controls the steering wheel and accelerator based on his/her own judgment so as to avoid interference between the vehicles 7 .

A non-compliant vehicle is a vehicle 7 that can transmit mobile information to the control device 10 through the wireless communication device 3 but cannot receive virtual traffic rules from the control device 10 .

Hereinafter, when it is not necessary to distinguish between compliant vehicles, semi-compliant vehicles, non-compliant vehicles, and non-connected vehicles, they are collectively referred to as "vehicle 7", and when they are described separately, they are referred to as compliant vehicles. 7A, semi-compliant vehicle 7B, non-compliant vehicle 7C, and non-connected vehicle 7D.

Further, the semi-compliant vehicle 7B, the non-compliant vehicle 7C, and the non-connected vehicle 7D are sometimes referred to as "manual vehicles" because the vehicle 7 is operated by the driver. On the other hand, since the compliant vehicle 7A is driven by the vehicle 7's own judgment and control, it may be referred to as an "automatic vehicle" in contrast to the manual vehicle.

The degree to which the movement of the vehicle 7 is realized along the route 6 instructed by the control device 10 by the control information or the degree to which the movement of the vehicle 7 is realized along the route 6 predicted by the control device 10 is referred to as the "observance degree". A vehicle 7 with a higher degree of compliance has a higher probability of moving along the route 6 instructed by the control device 10 or the route 6 predicted by the control device 10 . Since the movement of the vehicle 7 is restricted by the virtual traffic rules, the degree of observance may be referred to as the degree of observance of the virtual traffic rules or the degree of observance regarding movement.

Among the vehicles 7, the observance vehicle 7A has the highest observance degree, and the semi-observance vehicle 7B has the second highest observance degree. The compliance level of the non-compliant vehicle 7C is lower than that of the semi-compliant vehicle 7B, and the compliance level of the non-connected vehicle 7D is the lowest among the vehicles 7. That is, the types of vehicles 7 according to the present embodiment are classified according to the degree of observance of each vehicle 7 .

FIG. 2 is a diagram showing a functional configuration example of the control device of the observance vehicle 7A.

As shown in FIG. 2, the control device of the observance vehicle 7A includes functional units including a position estimation unit 21A, a state management unit 22A, a wireless communication unit 23A, a local route planning unit 24, and a control unit 25A, and a vehicle travel route map. 26.

The position estimation unit 21A estimates the position of the compliant vehicle 7A. Specifically, the position estimating unit 21A uses, for example, a laser range finder and a camera attached to the observance vehicle 7A to determine the shape of the surrounding environment in which the observance vehicle 7A travels and the movement path of the observance vehicle 7A, which is a moving body. Acquire external shape data from an external sensor that captures the situation on the lane 8, which is an example, and match the acquired external shape data with, for example, the environmental external shape data (self-position estimation map) prepared in advance. The position of the observant vehicle 7A is estimated using map matching that estimates the point as the current position of the observant vehicle 7A. There is no restriction on the method of estimating the position of the compliant vehicle 7A in the position estimating section 21A, and other estimation methods such as estimating the position of the compliant vehicle 7A using GPS (Global Positioning System), for example, may be used. The position estimation unit 21A notifies the state management unit 22A and the local route planning unit 24 of the estimated position of the compliant vehicle 7A as position information. 21 A of such position estimation parts are equivalent to the 1st position estimation part which concerns on this embodiment.

The state management unit 22A manages vehicle-specific information including information used to uniquely identify the compliant vehicle 7A, such as number information and vehicle body number. In addition, the state management unit 22A also detects the position of the vehicle 7 represented by the position information, as well as sensors (referred to as "inside sensors") that measure the attitude, speed, control details, etc. sensor values relating to the traffic environment are collected in chronological order and managed as the state of the observant vehicle 7A.

The state management unit 22A notifies the radio communication unit 23A and the local route planning unit 24 of the vehicle specific information and the state of the observance vehicle 7A as mobile body information at least one of regular and occasional timings.

The wireless communication unit 23A transmits the mobile object information received from the state management unit 22A to the control device 10 through the wireless communication device 3, and the control information received from the control device 10 through the wireless communication device 3 without delay for local route planning. Notify the department 24.

The local route planning unit 24 uses the position information received from the position estimation unit 21A, the mobile information received from the state management unit 22A, the control information received from the radio communication unit 23A, and the vehicle travel route map 26 to plan a local route. Plan and notify the control unit 25A.

The local route planning unit 24 complies with the obeying vehicle information contained in the mobile body information under the condition that the obeying vehicle 7A satisfies the virtual traffic rules contained in the control information and travels along the global route specified by the control information. The state of the roadway during travel is determined from the measurement data of the external sensor attached to the vehicle 7A, and the actual route is determined, such as which position on the roadway corresponding to the global route should actually be traveled. Then, the local route planning unit 24 sets the speed and attitude of the observance vehicle 7A at each time to follow in order for the vehicle 7 to travel along the determined actual route.

In this way, the actual route determined under the condition that the observant vehicle 7A travels along the global route specified by the control device 10 while satisfying the virtual traffic rules is called a "local route". , the control content for driving the observance vehicle 7A along the local route is added. Details of the global route will be described later.

The vehicle travel route map 26 includes map information representing the lanes 8 on which the observant vehicle 7A travels, and determines the local route and maps the observant vehicle 7A at each time for driving the observant vehicle 7A along the local route. It is used for setting the contents of control.

When the control unit 25A receives the local route from the local route planning unit 24, it controls the steering wheel, accelerator, brake, etc. of the observance vehicle 7A according to the control contents included in the local route, and causes the observance vehicle 7A to autonomously travel along the local route. Realize.

The control contents of the observance vehicle 7A, such as the operation amount of the steering wheel, accelerator, brake, etc. accompanying the control performed by the control unit 25A, are notified to the state management unit 22A through respective internal sensors that measure various control amounts, and the observance vehicle 7A is managed by the state management unit 22A.

Note that even if the observant vehicle 7A does not receive the global route from the control device 10, it is possible to determine the roadway conditions on which it is traveling from the measurement data of the external sensor, and to determine the actual route that satisfies the virtual traffic rules. However, as an example here, the observance vehicle 7A receives the global route from the control device 10 and determines the actual route, that is, the local route.

FIG. 3 is a diagram showing a functional configuration example of the control device of the semi-compliant vehicle 7B.

As shown in FIG. 3, the control device of the semi-compliant vehicle 7B includes a position estimation unit 21B, a state management unit 22B, a wireless communication unit 23B, a control unit 25B, and a virtual traffic rule transmission unit 27.

Since the driver determines the actual route of the quasi-compliant vehicle 7B, unlike the compliant vehicle 7A, the local route planning unit 24 and the vehicle travel route map 26 do not exist, and instead, the virtual traffic rule transmission unit 27 exists. .

The position estimation unit 21B estimates the position of the quasi-compliant vehicle 7B by the same method as the position estimation unit 21A in the compliant vehicle 7A, and notifies the state management unit 22B of the estimated position of the quasi-compliant vehicle 7B as position information.

The state management unit 22B, like the state management unit 22A in the observance vehicle 7A, collects and manages the moving body information of the semi-observance vehicle 7B. In the case of the semi-observant vehicle 7B, since the driver controls the semi-observant vehicle 7B, the state of the semi-observant vehicle 7B includes the operation amount of the steering wheel, accelerator, brake, etc. operated by the driver.

In addition, the state management unit 22B notifies the wireless communication unit 23B of the mobile unit information to be managed at least one of regular and occasional timings.

The wireless communication unit 23B transmits the mobile object information received from the state management unit 22B to the control device 10 through the wireless communication device 3, and also transmits the control information received from the control device 10 through the wireless communication device 3 to the virtual traffic rule without delay. The transmission unit 27 is notified.

The virtual traffic rule transmission unit 27 transmits the virtual traffic rule included in the control information received from the control device 10 through the wireless communication unit 23B to the driver. As long as the driver of the quasi-observant vehicle 7B can comprehend the virtual traffic rules, there are no restrictions on the method of transmitting the virtual traffic rules in the virtual traffic rules transmitting unit 27. Any method of transmission may be used, or a plurality of transmission methods may be combined to transmit the virtual traffic rules to the driver.

Incidentally, the quasi-compliant vehicle 7B may receive the global route from the control device 10 and transmit the global route together with the virtual traffic rules to the driver by the virtual traffic rule transmission unit 27 to assist the driver in driving. As an example, the semi-compliant vehicle 7</b>B does not receive the global route from the control device 10 .

FIG. 4 is a diagram showing a functional configuration example of the control device of the non-compliant vehicle 7C.

As shown in FIG. 4, the control device of the non-compliant vehicle 7C includes a position estimation section 21C, a state management section 22C, a wireless communication section 23C, and a control section 25C.

Since the non-observing vehicle 7C cannot receive the virtual traffic rules from the control device 10 or does not have the function of utilizing the virtual traffic rules even if it receives the virtual traffic rules, It becomes the structure from which the rule transmission part 27 was removed.

The position estimator 21C estimates the position of the non-compliant vehicle 7C by the same method as the position estimator 21A for the compliant vehicle 7A, and notifies the state management unit 22C of the estimated position of the non-compliant vehicle 7C as position information.

The state management unit 22C collects and manages moving body information of the non-observance vehicle 7C in the same manner as the state management unit 22B in the quasi-observance vehicle 7B. In addition, the state management unit 22C notifies the wireless communication unit 23C of the mobile unit information to be managed at least one of regular and occasional timings.

The radio communication unit 23C transmits the mobile unit information received from the state management unit 22C to the control device 10 through the radio communication device 3 .

The non-observing vehicle 7C may receive the global route from the control device 10 and transmit the received global route to the driver to assist the driver in driving. 7C will be described as not receiving the global route from the control device 10 .

If the non-observing vehicle 7C cannot receive the virtual traffic rules from the control device 10, or if it does not have the function of utilizing the virtual traffic rules even if it receives the virtual traffic rules, it has some automatic driving function. However, as an example here, it is assumed that the driver is manually driving the non-compliant vehicle 7C.

FIG. 5 is a diagram summarizing the differences in driving functions of the observance vehicle 7A, the semi-observance vehicle 7B, the non-observance vehicle 7C, and the non-connected vehicle 7D.

In FIG. 5, "O" indicates that the corresponding operating function is provided, and "X" indicates that the corresponding operating function is not provided. "O/X" indicates that it does not matter whether the corresponding operation function is provided or not, and "-" indicates that the operation function is unnecessary in principle.

Hereinafter, for convenience of explanation, when it is necessary to distinguish between the mobile information, the mobile information of the observant vehicle 7A will be referred to as "mobile information A", and the mobile information of the quasi-compliant vehicle 7B will be referred to as "mobile information B". , and the mobile body information of the non-compliant vehicle 7C will be referred to as "mobile body information C".

On the other hand, FIG. 6 is a diagram showing a functional configuration example of the collection device 9 and the control device 10. As shown in FIG. As shown in FIG. 6, the collection device 9 includes an infrastructure sensor 9A and a state detector 9B.

The infrastructure sensor 9A is, for example, a photographing unit that optically photographs, a UWB radar, or the like, and photographs or measures the unconnected vehicle 7D and the traffic environment around the unconnected vehicle 7D. The infrastructure sensor 9A notifies the state detector 9B of collected data relating to the traveling state and traffic environment of the unconnected vehicle 7D collected by photography or measurement. For example, the infrastructure sensor 9A shoots or measures the traffic environment at predetermined intervals. may be performed. As an example, the infrastructure sensor 9A is assumed to be a camera that constantly captures images at predetermined intervals. If the image capturing interval is set short, the image captured by the infrasensor 9A will be a moving image. Such infrastructure sensor 9A corresponds to an acquisition device that acquires the state of vehicle 7 .

Upon receiving the collected data from the infrastructure sensor 9A, the state detection unit 9B extracts the vehicle portion of the unconnected vehicle 7D from the image captured by the infrastructure sensor 9A by known image processing, and determines the position of the unconnected vehicle 7D in the image. The position of the unconnected vehicle 7D on the controlled travel route map 19 is calculated by performing coordinate conversion from the position of the camera coordinates in the infrastructure sensor 9A to the position of the map coordinates representing the actual shape of the road. Once the position of the non-connected vehicle 7D on the control travel route map 19 is determined, for example, from the positional deviation of the non-connected vehicle 7D obtained from the two images and the time difference between the two images, the non-connected vehicle A speed of 7D is obtained. Also, it is possible to track how the non-connected vehicle 7D is moving from changes in the position of the non-connected vehicle 7D along the time series.

In this way, the state detection unit 9B detects mobile information of the non-connected vehicle 7D, such as the position, posture, and speed of the non-connected vehicle 7D, and transmits the information to the control device 10 without delay through the communication network 5. Henceforth, the mobile body information of non-connected vehicle 7D may be called "mobile body information D."

The collection device 9 does not necessarily have the infrasensor 9A and the state detection unit 9B, and may have only the infrasensor 9A, for example. In this case, a state detection device (not shown) having a state detection unit 9B is connected to the communication network 5, and the state detection device (not shown) receives each collection data notified from each collection device 9, and the unconnected vehicle 7D What is necessary is just to detect mobile body information. Since the state detection unit 9B is not required for each collection device 9, the collection device 9 can be made smaller than when the state detection unit 9B is included in the collection device 9, and the collection device 9 can be easily attached.

Similarly to the wireless communication device 3, the collection device 9 is installed in a range where mobile information of the non-connected vehicle 7D can be obtained, such as a sidewalk provided parallel to the roadway, a median strip of the roadway, and signal equipment. be done. The photographing direction of the infrasensor 9A in the collection device 9 and the range (angle of view) of the photographed image may be fixed or variable.

Next, functions of the control device 10 will be described.

As shown in FIG. 6, the control device 10 includes a destination setting unit 11, a communication unit 12, a global route planning unit 13, a travel route prediction unit 14, an interference point identification unit 15, an anomaly detection unit 16, and a virtual traffic rule generation unit. Each functional part of the part 18 and the control driving|running route map 19 are included.

The communication unit 12 receives moving body information from each of the observant vehicle 7A, the semi-observant vehicle 7B, and the non-observant vehicle 7C through the wireless communication device 3, and generates a virtual traffic rule generated by a virtual traffic rule generation unit 18, which will be described later. The included control information is transmitted to each of the compliant vehicles 7A and semi-compliant vehicles 7B designated as destinations of the control information. The communication unit 12 is an example of a receiving unit that receives mobile information.

Since the compliant vehicle 7A is a vehicle 7 that travels along the global route designated by the control device 10, the control device 10 needs to plan the global route.

Therefore, the destination setting unit 11 receives the destination to which the observant vehicle 7A is going and the vehicle-specific information of the observant vehicle 7A, associates the vehicle-specific information of the observant vehicle 7A with the destination, and outputs the information to the global route planning unit 13. Notice.

The destination and vehicle-specific information of the compliance vehicle 7A may be set in the destination setting unit 11 by the operator of the control device 10. However, the destination is set by receiving the mobile body information A including the destination from the compliance vehicle 7A. The unit 11 may associate the vehicle-specific information and the destination included in the received moving body information A and notify the global route planning unit 13 of them.

The global route planning unit 13 plans a global route using the destination of the observant vehicle 7A received from the destination setting unit 11, the moving body information A of the observant vehicle 7A received from the communication unit 12, and the control travel route map 19. do. The global route planning unit 13 notifies the interference point identifying unit 15 of the planned global route.

FIG. 7 is a diagram showing an example of the contents of the control driving route map 19 used for planning global routes. The controlled driving route map 19 is a road structure database represented by a set of route information for each section, with sections where the traffic rules do not change as the minimum unit.

Sections constituting the control driving route map 19 (in the example of FIG. 7, the section from the point K1 to the point K2, the section from the point K2 to the point K3, the section from the point K3 to the point K7, the section from the point K2 to the point K5 , the section from point K4 to point K5, and the section from point K5 to point K6), the speed limit, the number of lanes, the curvature, the width, the slope, and the lanes in the section intersect In this case, traffic rule information such as lane priority, such as whether the lane is a priority lane or a non-priority lane, is set.

Further, each section constituting the control travel route map 19 represents the route 6 of the vehicle 7 by a series of waypoints, which is a set of virtually set waypoints. A waypoint is one of indices representing the position of the vehicle 7 on the route 6. Since a waypoint ID is uniquely set for each waypoint, the vehicle 7 on the route 6 can be determined from the waypoint ID. is located.

Note that the control travel route map 19 includes interference point information that defines the interference point X in advance.

FIG. 8 is a diagram showing an example of an interference point X. FIG. Interference point X includes, for example, an intersection where priority lane 8A and non-priority lane 8B intersect, as shown in FIG. 8A, and a priority lane shown in FIG. Includes a junction where 8A and non-priority lane 8B join. Since the interference point X is a place where the vehicles 7 approach each other within the interference range, the interference point X is not necessarily represented by a point, but is represented by an area having a certain size. There is also That is, the interference point X is an example of the interference area.

FIG. 9 is a diagram showing an example of interference point information. The interference point information includes, for example, an interference point ID for uniquely identifying each interference point X, the position of the interference point X, The waypoint ID for uniquely identifying the corresponding waypoint and the position and waypoint ID of the waypoint immediately before the interfering point X are assigned to the priority lane 8A and the non-priority lane 8B that share the same interference point X. It is the information specified for each of

The vehicle travel route map 26 for the observance vehicle 7A described above is configured with the same information as the control travel route map 19, but the vehicle travel route map 26 is information necessary for the observance vehicle 7A to actually travel, and Information that is not included in the control travel route map 19 may be included.

Note that the control travel route map 19 does not necessarily include the interference point information. good. Also, the control travel route map 19 does not necessarily have to be included in the control device 10 , and the control device 10 may acquire the control travel route map 19 from an external device different from the control device 10 through the communication unit 12 .

The global route planning unit 13 shown in FIG. Based on the route map 19, a global route, which is the route 6 from the current position of the observant vehicle 7A to the destination, is searched on the control travel route map 19 for each observant vehicle 7A. In other words, the global route is expressed as a route 6 that connects the series of waypoints in each section forming the control travel route map 19 .

FIG. 10 is a diagram showing an example of global routes of vehicles P and Q, both of which are observant vehicles 7A. One route 6P connecting the current position of the vehicle P and the destination of the vehicle P is selected from among the plurality of routes 6 as the global route of the vehicle P, and one route connecting the current position of the vehicle Q and the destination of the vehicle Q is selected. 6Q is selected as vehicle Q's global route.

Any search method may be used for searching the global route of the observance vehicle 7A in the global route planning unit 13. For example, a known search method such as Dijkstra search is used.

Since the destination of the compliance vehicle 7A is determined in advance, the global route planning unit 13 can plan the global route of the compliance vehicle 7A mainly by the control device 10 . Since the observance vehicle 7A travels on the local route determined based on the global route, the control device 10 can know the route 6 of the observance vehicle 7A from the global route planned by itself.

On the other hand, in the case of the semi-observant vehicle 7B, since the driver determines the route 6, the destination may not be determined in advance. Also, if the non-observance vehicle 7C is also driven by a driver, the driver determines the route 6 in the same way as the semi-observance vehicle 7B, so the destination may not be determined in advance. Furthermore, since the non-connected vehicle 7D cannot perform data communication with the control device 10, the control device 10 has no means of acquiring the destination to which the non-connected vehicle 7D is going.

Therefore, the global route planning unit 13 of the control device 10 cannot plan global routes for the semi-compliant vehicle 7B, the non-compliant vehicle 7C, and the non-connected vehicle 7D.

Instead, the control device 10 uses the travel route prediction unit 14 to predict the travel routes of the semi-compliant vehicle 7B, the non-compliant vehicle 7C, and the non-connected vehicle 7D.

Specifically, the travel route prediction unit 14 calculates the moving object information B, the moving object information C, and the moving object information D of each of the semi-compliant vehicle 7B, the non-compliant vehicle 7C, and the non-connected vehicle 7D, that is, the manual vehicle. It predicts the driving route from the position information and speed contained in . For example, the travel route prediction unit 14 assumes that the manual vehicle travels from the current position of the manual vehicle to a specific distance (for example, 100 m) at a speed indicated by the mobile object information. is notified to the interference point identification unit 15 as the traveling route of the manual vehicle.

However, since the destination of the manually operated vehicle is unknown, when the lane 8 diverges, the travel route prediction unit 14 travels a combination of all the routes 6 branching each time the lane 8 diverges, unlike the global route. Treat as a route.

FIG. 11 is a diagram showing an example of a travel route when the vehicle P is a manually operated vehicle. As shown in FIG. 11, when the destination of the vehicle P branches into two, the traveling route prediction unit 14 does not know in which direction the vehicle P branches at the branch point. and the route 6P-2 are both the travel route of the vehicle P.

As already explained, the mobile information B of the quasi-compliant vehicle 7B and the mobile information C of the non-compliant vehicle 7C are acquired through the wireless communication device 3, and the mobile information D of the non-connected vehicle 7D is acquired through the collection device 9. is obtained.

Images captured by the collection device 9 may include vehicles 7 of types other than the non-connected vehicle 7D. In this case, it may not be possible to distinguish which mobile object information acquired from the collection device 9 is the mobile object information D of the non-connected vehicle 7D. Therefore, the travel route prediction unit 14, for example, any of the position information included in the mobile object information A, the mobile object information B, and the mobile object information C acquired through the communication unit 12, and the position information of the vehicle 7 acquired from the collection device 9 When the vehicle 7 represented by the mobile object information notified from the collection device 9 is approaching within a predetermined range where the position information included in the mobile object information is recognized to match, the vehicle 7 is notified by the communication unit 12 It may be determined that the vehicle is any one of the observant vehicle 7A, the semi-observant vehicle 7B, and the non-observant vehicle 7C represented by the acquired mobile object information, and distinguished from the mobile object information D of the non-connected vehicle 7D.

The interference point identification unit 15 receives the global route of the compliant vehicle 7A from the global route planning unit 13, and the traveling routes of the semi-compliant vehicle 7B, the non-compliant vehicle 7C, and the non-connected vehicle 7D predicted by the traveling route prediction unit 14. , the interference point X on each path 6 is identified. In addition, the interference point identification unit 15 notifies the route 6 of each vehicle 7 and the interference point X on each route 6 to the virtual traffic rule generation unit 18 .

The abnormality detection unit 16 receives moving body information of each vehicle 7 from the communication unit 12 and the collecting device 9, and detects an abnormality related to movement of the approaching vehicle based on changes in the state related to movement of the approaching vehicle represented by the received moving body information. Detect whether or not it has occurred.

Here, the term "approaching vehicle" refers to a vehicle 7 that has a possibility of approaching the specific vehicle 7 within a predetermined range at the interference point X, among the vehicles 7 that have received moving body information for the specific vehicle 7. It's about. Further, the “abnormality related to the movement” of the approaching vehicle means that the approaching vehicle makes an unexpected movement that is not predicted by the control device 10 .

If it is found that a vehicle 7 approaching a specific vehicle 7 has an abnormality related to movement, the control device 10 controls the movement of the specific vehicle 7 so as not to interfere with the vehicle 7 approaching at the interference point X. By transmitting a virtual traffic rule with stronger restrictions, it is possible to prevent a specific vehicle 7 from interfering with an approaching vehicle at the interference point X.

However, when the specific vehicle 7 is a manual vehicle, the driver does not necessarily drive the manual vehicle according to the virtual traffic rules, and in particular if the specific vehicle 7 is a non-connected vehicle 7D, the driver receives the virtual traffic rules. I can't. That is, the specific vehicle 7 is preferably a compliant vehicle 7A that is driven according to the virtual traffic rules. Therefore, the abnormality detection unit 16 according to the present embodiment recognizes a manual vehicle approaching the observance vehicle 7A as an approaching vehicle, and detects whether or not an abnormality related to the movement of the manual vehicle has occurred.

The abnormality detection unit 16 notifies the virtual traffic rule generation unit 18 of the determination result indicating whether or not an abnormality related to the movement of the approaching vehicle has occurred.

In addition, hereinafter, the abnormality related to the movement of the approaching vehicle is simply referred to as the "abnormality" of the approaching vehicle. A method of detecting an abnormality in an approaching vehicle will be described later in detail. In FIG. 6, in order to clearly show the notification destination of the mobile body information D from the collection device 9, the mobile body information D is notified from the collection device 9 to the travel route prediction unit 14 and the abnormality detection unit 16. is notified to the travel route prediction unit 14 and the abnormality detection unit 16 via the communication unit 12 .

The virtual traffic rule generation unit 18 generates the moving body information of each vehicle 7 received from the communication unit 12, the route 6 of each vehicle 7 received from the interference point identification unit 15, that is, the global route and the travel route, and each route 6 , and the determination result regarding the abnormality of the approaching vehicle received from the abnormality detection unit 16, virtual traffic rules for the compliant vehicle 7A and the semi-compliant vehicle 7B are generated. The virtual traffic rule generator 18 corresponds to the generator according to this embodiment.

Specifically, the virtual traffic rule generation unit 18 first uses the moving body information of each vehicle 7, the route 6 of each vehicle 7, and the interference point X on each route 6 to determine the interference on the global route of the observant vehicle 7A. For the observance vehicle 7A and other vehicles 7 that interfere with each other at the point X, the interference time t at each interference point X is estimated, and the passage priority is observed so that the interference between the vehicles 7 is avoided at each interference point X. Defined for vehicle 7A.

Similarly, the virtual traffic rule generation unit 18 uses the moving body information of each vehicle 7, the route 6 of each vehicle 7, and the interference point X on each route 6 to determine the interference point X on the travel route of the quasi-compliant vehicle 7B. estimating the interference time t at each interference point X for the quasi-compliant vehicle 7B and the other vehicle 7 that interfere with each other, and quasi-complying the passage priority order to avoid interference between the vehicles 7 at each interference point X Defined for vehicle 7B.

After that, the virtual traffic rule generation unit 18 refers to the determination result regarding the abnormality of the approaching vehicle, and determines whether the approaching vehicle, which may interfere with the observance vehicle 7A at the interference point X on the global route of the observance vehicle 7A, has an abnormality. is detected, based on the design concept of fail-safe, it will be less likely to interfere with the approaching vehicle at the interference point X than the virtual traffic rule generated before referring to the judgment result regarding the abnormality of the approaching vehicle. Generate enhanced virtual traffic rules with stronger restrictions on movement. The enhanced virtual traffic rule corresponds to the enhanced constraint condition according to this embodiment.

The virtual traffic rule generation unit 18 generates control information including the global route and the virtual traffic rules for the observant vehicle 7A, and generates control information including the virtual traffic rules for the quasi-observant vehicle 7B, The communication unit 12 is requested to transmit the control information associated with each of the compliant vehicles 7A and semi-compliant vehicles 7B. As a result, the control information associated with each of the observance vehicle 7A and the semi-observance vehicle 7B is transmitted from the wireless communication device 3 to the corresponding observance vehicle 7A and semi-observance vehicle 7B.

A detailed method of generating virtual traffic rules and enhanced virtual traffic rules will be described later.

Next, a configuration example of main parts of each electric system in the control device 10 and the control device of the vehicle 7 will be described.

FIG. 12 is a diagram showing a configuration example of a main part of an electric system in the control device 10. As shown in FIG. As shown in FIG. 12, the control device 10 is configured using a computer 30, for example.

The computer 30 includes a CPU (Central Processing Unit) 31, which is an example of a processor responsible for processing each functional unit related to the control device 10 shown in FIG. Read Only Memory) 32 , RAM (Random Access Memory) 33 used as a temporary work area for CPU 31 , nonvolatile memory 34 , and input/output interface (I/O) 35 . A CPU 31 , ROM 32 , RAM 33 , nonvolatile memory 34 and I/O 35 are connected via a bus 36 .

The nonvolatile memory 34 is an example of a storage device that maintains stored data even when the power supplied to the nonvolatile memory 34 is interrupted. For example, a semiconductor memory is used, but a hard disk may be used. In addition, the nonvolatile memory 34 does not necessarily have to be built in the computer 30. For example, a portable storage medium such as a USB (Universal Serial Bus) memory or a memory card that can be attached to and detached from the computer 30 may be used. .

A communication unit 37, an input unit 38, and a display unit 39 are connected to the I/O 35, for example.

The communication unit 37 is connected to the communication network 5 and has a communication protocol for data communication between the wireless communication device 3 , the collection device 9 , and an external device connected to the communication network 5 .

The input unit 38 is a unit that receives instructions from the operator of the control device 10 and notifies the CPU 31 of them, and uses buttons, a touch panel, a keyboard, a mouse, and the like, for example. A microphone may be used as the input unit 38 if the instructions are given by voice.

The display unit 39 is a device that displays information processed by the CPU 31, and uses, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, a projector, or the like.

The display unit 39 may not be connected to the I/O 35 in situations where information display is not required, such as when the control device 10 is installed in an unmanned data center or the like. Other units, such as a printer unit, may also be connected to I/O 35 as needed.

On the other hand, FIG. 13 is a diagram showing a main configuration example of an electrical system in a control device for a vehicle 7 other than a non-connected vehicle 7D. The control device of the vehicle 7 is also configured using a computer 40, for example, like the control device 10. FIG.

The computer 40 is a CPU 41, which is an example of a processor responsible for the processing of each functional unit related to the control device for each type of vehicle 7 shown in FIGS. A ROM 42 that stores a control program that functions as a control program, a RAM 43 that is used as a temporary work area for the CPU 41, a nonvolatile memory 44, and an I/O 45. A CPU 41 , ROM 42 , RAM 43 , nonvolatile memory 44 and I/O 45 are connected via a bus 46 .

A communication unit 47, an input unit 48, a display unit 49, an internal sensor 51, an external sensor 52, and a traveling device 53 are connected to the I/O 45, for example.

The communication unit 47 has a communication protocol for data communication with the wireless communication device 3 .

The input unit 48 is a unit that receives instructions from the driver of the vehicle 7 and notifies the CPU 41 of them, and uses, for example, buttons, a touch panel, and a pointing device. A microphone may be used as the input unit 48 when the instructions are given by voice.

The display unit 49 is a device that displays information processed by the CPU 41, and uses, for example, a liquid crystal display, an organic EL display, a projector, or the like.

The travel device 53 is a device that adjusts the operation amount of an operating device that affects the travel state of the vehicle 7, such as a steering wheel, accelerator, and brake, and notifies the CPU 41 of the operation amount as the state of the vehicle 7. In the case of the observance vehicle 7A, the operation amount is controlled by the CPU 41, and in the case of the semi-observance vehicle 7B, the operation amount is controlled by the driver. Since the non-observing vehicle 7C may or may not have an automatic driving function, the operation amount of the operation device may be controlled by the CPU 41 according to the driving mode, or may be controlled by the driver. However, as already explained, since the example in which the driver manually drives the non-observing vehicle 7C is explained, the amount of operation of the operation device of the non-observing vehicle 7C in this embodiment depends on the driver. controlled.

It should be noted that the I/O 45 may be connected to a speaker that notifies the content of the control information received from the control device 10 by voice, or a device that notifies the content by touch. If the vehicle 7 is a semi-compliant vehicle 7B, at least one device is connected to the I/O 45 for communicating control information to the driver. The functions of the internal sensor 51 and the external sensor 52 have already been described. sometimes not.

Next, the operation of the control device 10 in the control system 1 will be described in detail.

FIG. 14 is a flow chart showing an example of the flow of control processing executed by the CPU 31 of the control device 10 when mobile body information is received from each vehicle 7 .

A control program that defines control processing is stored in advance in the ROM 32 of the control device 10, for example. The CPU 31 of the control device 10 reads a control program stored in the ROM 32 and executes control processing.

For convenience of explanation, the operation of the control device 10 will be described below for the case where the vehicles 7 to be controlled by the control system 1 are the vehicle P and the vehicle Q. Even if there is, the same processing may be performed for each vehicle 7 .

In step S10, the CPU 31 determines the types of the vehicle P and the vehicle Q based on the driving function of the vehicle 7 notified by the mobile body information. For convenience of explanation, it is assumed that the vehicle P is a manual vehicle and the vehicle Q is an automatic vehicle.

In step S20, the CPU 31 generates a global route for the vehicle Q based on the location information of the vehicle Q and the destination of the vehicle Q included in the moving body information received from the vehicle Q, which is an automobile.

In step S30, the CPU 31 generates the travel route of the vehicle P based on the position information of the vehicle P and the speed of the vehicle P included in the moving body information received from the vehicle P, which is a manually operated vehicle.

Note that when the speed of the manual vehicle slows down at a point before the branch point, it is more likely that the vehicle will turn around than if it goes straight through the branch point. You may

In step S40, the CPU 31 determines a point where the waypoint indicating the route 6 of the vehicle P approaches within the range of interference from the waypoint indicating the route 6 of the vehicle Q, that is, the point of interference X between the vehicle P and the vehicle Q. Identify. Specifically, the CPU 31 determines the global route of vehicle Q acquired in step S20 (simply referred to as "route 6 of vehicle Q") and the travel route of vehicle P acquired in step S30 (simply referred to as "route 6 of vehicle P"). ) intersect or merge, or in an area where the route 6 of the vehicle P and the route 6 of the vehicle Q run side by side within the interference range. Identify as point X.

Note that when the interference point information as shown in FIG. 9 can be referred to, the CPU 31 sets the route 6 of each of the vehicles P and Q to the same route set as the interference point X in the interference point information. If the point is included, the point represented by the waypoint may be the point of interference X between the vehicle P and the vehicle Q.

In step S50, the CPU 31 calculates an interference time t between the vehicle P and the vehicle Q for each collision point X between the vehicle P and the vehicle Q specified in step S40.

The interference time t is the distance along the route 6 from the current positions of the vehicles P and Q to the interference point X (hereinafter referred to as "the distance to the interference point X"), and the distance between the vehicle P and the vehicle Q. It can be calculated from the velocity. However, because the vehicle P and the vehicle Q cannot travel at a constant speed due to traffic conditions and the like, the speed of the vehicle P and the vehicle Q actually fluctuates.

For example, "v" is the respective velocities of vehicle P and vehicle Q, "l" is the distance to the point of interference X, and "α" and "β" are positive margins corresponding to the velocity changes of vehicle P and vehicle Q. , "ε" is a positive constant closer to 0 than other values, the interference time t at the interference point X considering the amount of variation in the speed of the vehicle P and the vehicle Q is calculated by equation (1).

Note that t _begin is the time when vehicle P and vehicle Q pass the interfering point X earliest, and t _end is the time when vehicle P and vehicle Q pass the interfering point X latest. Thus, the interference time t at each interference point X of the vehicle P and the vehicle Q is represented by a period of _tbegin or more and less than _tend . Hereinafter, the interference time t of the vehicle P at the interference point X will be referred to as "interference time t ^p ", and the interference time t of the vehicle Q at the interference point X will be referred to as "interference time t ^q ". The method of calculating the interference time t is not limited to the above, and prediction using a driver model or vehicle behavior model may be used.

FIG. 15 is a diagram showing an example of an interference time t ^p of a vehicle P and an interference time t ^q of a vehicle Q with respect to a specific interference point X. In FIG.

In this manner, the CPU 31 calculates the vehicle P interference time t ^p and the vehicle Q interference time t ^q for each interference point X. FIG.

In step S60, the CPU 31 sets passage priority at each interference point X of the vehicle P and the vehicle Q identified in step S40.

A vehicle 7 with a higher passing priority, that is, with a priority passing priority, can pass through the interference point X preferentially, and has a lower passing priority, that is, with a non-prioritizing passing priority. A restriction is set so that the vehicle 7 whose passage priority is set not to enter the interference point X during the interference time t at the interference point X of the vehicle 7 whose passage priority is set to priority.

In this way, by associating the passage priority and the interference time t at each interference point X and restricting the time and place where the vehicle P and the vehicle Q can run, the vehicle P and the vehicle Q at each interference point X interference is avoided.

Correspondence between the passage priority at each interference point X and the interference time t of the vehicle 7 whose passage priority is set to priority, that is, the correspondence to the interference point X for the vehicle 7 whose passage priority is set to non-priority No-entry time becomes a virtual traffic rule. The virtual traffic rule restricts the time and place where the vehicle 7 can travel by using the interference time t and the position information of the interference point X so that the vehicle 7 does not interfere with another vehicle 7 at the interference point X. An example of spatial constraints.

Hereinafter, the state indicating whether the priority order of passage at the interference point X is set to priority or non-priority is represented by "s _x ". The state s _x at the point of interference X takes a value of “0” or “1”, and the CPU 31 sets the state s x =0 when the vehicle P has priority at the point of interference X, and sets the state s _x =0 when the vehicle Q has priority. , set the state s _x =1.

FIG. 16 is a schematic diagram that schematically illustrates the virtual traffic rules at the interference point X in which the state s _x =0 is set.

In this case, if the interference time t ^p of the vehicle P is expressed by a period from t ^p _begin to less than t ^p _end , the vehicle P gives priority to the interference point X over the vehicle Q during the interference time t ^p . can be passed through However, if the vehicle Q enters the interference point X during the interference time tp, it may interfere with the vehicle ^P , so it is necessary not to enter the interference point X during the interference time ^tp .

On the other hand, FIG. 17 is a schematic diagram that schematically illustrates the virtual traffic rules at the interference point X in which the state s _x =1 is set.

In this case, if the vehicle Q's interference time t ^q is expressed by a period from t ^q _begin to less than t ^q _end , the vehicle Q gives priority to the interference point X over the vehicle P during the interference time t ^q . can be passed through However, the vehicle P may interfere with the vehicle Q if it enters the interference point X during the interference time t ^q , so it is necessary not to enter the interference point X during the interference time t ^q .

In other words, the vehicle 7 whose passage priority at the interference point X is set to non-priority is prevented from entering the interference point X during the interference time t of the vehicle 7 whose passage priority is set to priority at the same interference point X. Then, it is possible to avoid interference between the vehicles 7 and reach the destination.

In this way, the virtual traffic rule is a combination of the state s _x at the interference point X and the entry prohibition time at the interference point X (that is, the interference time t of the vehicle 7 whose passage priority is set to be higher at the interference point X). is represented by information set for each interference point X.

How to determine the transit priority of vehicle P and vehicle Q at each interference point X is performed by evaluating the objective function C(S).

The CPU 31 determines the passing priority of the vehicle P and the vehicle Q using an objective function C(S) pre-stored in the non-volatile memory 34, for example. The objective function C(S) is a function representing the traffic cost generated by the movement of the vehicle 7 . The traffic cost is not only the cost for each vehicle 7 but also the cost that the traveling of the vehicle 7 affects the traveling of other vehicles 7 , and expresses the traffic efficiency of the vehicle 7 .

Therefore, the CPU 31 sets the state vector S that minimizes the objective function C(S). If the objective function C(S) is minimized, the traffic cost is also minimized, so the traffic efficiency in the area controlled by the control system 1 is improved.

Note that "S" in the objective function C(S) is defined as A set of states s _x is expressed as a vector and is called a "state vector". The state vector S is represented by S=(s ₁ , s ₂ , . . . s _i , . . . , s _N ).

Hereinafter, a determination method for determining the passage priority order of the vehicle P and the vehicle Q using the objective function C(S) will be described.

The objective function C(S) is represented, for example, by Equation (2).

Here, C _single (S) is a cost that characterizes traffic control for vehicle P and vehicle Q that pass through each interference point X. C _consistency (S) is the cost related to the temporal consistency of transit priorities set for vehicle P and vehicle Q; C _multi (S) is the cost of consistency and interaction between interference points X; ω _c and ω _m are weights for adjusting the degree of influence of C _consistency (S) and C _multi (S) on objective function C(S).

In equation (2), C _single (S) is defined for vehicles 7 (in this case, vehicle P and vehicle Q) passing through each interference point X, and is expressed by equations (3) and (4). be.

In other words, C _single (S) is set for each interference point X when the vehicle 7 with the shortest interference time t at the interference point X, that is, the vehicle 7 with the shortest expected passage time at the interference point X is given priority. take a value that is low.

The CPU 31 determines the priority of the lane 8 in which the vehicle 7 travels, the length and traffic density of the queue in the lane 8 in which the vehicle 7 travels, the time the vehicle 7 has waited so far, and the priority of the vehicle 7 itself such as an emergency vehicle. C _single (S) may be weighted taking into account the degree. For example, according to the equation (4), the vehicle P's interference time t ^p is later than the vehicle Q's interference time t ^q even though the vehicle P's passage priority is set to be high at a certain interference point _Xi . In this case, C _single (s _{i ;} t ^p , t ^q ) is set to "1 ^" _. t ^q ) may be weighted to a value greater than one.

In other words, when the priority passage order is set to be non-prioritized in a situation where it is originally considered better to set the priority passage order to priority, or when it is better to set the priority passage order to be non-prioritized. In a conceivable situation, if priority is given to priority passage, the CPU 31 may perform weighting so that C _single (S) is large.

In the equation (2), C _consistency (S), which represents the cost related to the temporal consistency of the transit priority set for the vehicle 7, is represented by the equations (5) and (6).

The state s _x is set to an optimum value for each predetermined time unit (for example, 1 second) along the time series with reference to the mobile body information at that time, but s _x ^(T) is the time The state s _x set at time T is represented, and s _x ^(T−1) represents the state s _x set at time T−1 one unit before time T.

The vehicle 7 can travel in the lane 8 efficiently if priority and non-priority indicating the order of passing priority for the same vehicle 7 are not frequently switched. Therefore, C _consistency (S) is set to be small if the values of the temporally adjacent states s _x repeatedly calculated along the time series are the same, and large if they are different. Take.

Here, as an example, the cost related to the consistency when the state s _x ^(T) is set is calculated from the relationship with the state s _x ^(T−1) set one unit earlier. A cost related to consistency when the state s _x ^(T) is set may be calculated from the relationship with the state s _x at each time up to M units (M is a positive integer of 2 or more). Specifically, the CPU 31 calculates the number of times priority and non-priority are switched in the period from time T to M units before each interference point X, and the number of times priority and non-priority are switched at each interference point X increases. C _consistency (S) may be calculated such that the value of C _consistency (S) increases as

In equation (2), C _multi (S) representing the cost related to consistency and interaction between interference points X is represented by equations (7) and (8).

Here, w _i,j is a value representing the degree of cooperation between two interference points X _i and X _j and is set to w _i,j ≧0. The degree of cooperation of the interfering point X indicates the degree to which it is considered better to link the passage priority of one interfering point X with the passage priority of the other interfering point X and set them to the same state.

wi _,j = 0 indicates that the interference points Xi, _Xj do not need to cooperate, and as _{wi ,j increases} , it is better to treat the two interference points Xi, _Xj cooperatively. represents

FIG. 18 is a diagram showing an example in which a plurality of interference points X should be coordinated and handled. As shown in FIG. 18, when a vehicle P traveling on the non-priority lane 8B crosses the one-way priority lane 8A on which another vehicle 7 is traveling and goes straight ahead, the vehicle P is located at the interference _point X1 and the interference point X1. _Need to go through X2.

_In such a situation, if the state s1 at the interference point X1 for the vehicle P is set to " ₀ " and the state _s2 at the interference point X2 is set to " ₁ ", the other vehicle ₇ at the interference point X2 In order to avoid interfering with the vehicle P, the vehicle P may stop on the priority lane 8A. In this case, the vehicle P obstructs the flow of vehicles 7 traveling on the priority lane 8A.

Therefore, when there are a plurality of interference points X at locations where it is better not to stop until the vehicle 7 finishes passing once it starts passing, the state s _x of each of the interference points X is set to the same value so that the interference occurs. It is better to coordinate the points X with each other.

The w _i,j for each interference point X combination is pre-computed and stored, for example, in non-volatile memory 34 . The CPU 31 may refer to w _i,j stored in the nonvolatile memory 34 and calculate C _multi (S).

Here, as an example, the calculation of C _multi (S) based on the degree of cooperation of two interfering points X has been described, but the degree of cooperation of three or more interfering points X is used to calculate C _multi (S). good too.

In this way, the objective function C(S) is the content of traffic control for the vehicle 7, the consistency in the time direction of the passage priority at the interfering point X, the cost reflecting the consistency and interaction between the interfering points X, and the lane 8 However, the objective function C(S) used by the CPU 31 to determine the passing priority of each vehicle 7 at the interfering point X is not limited to the objective function C(S) described above. An objective function C(S) reflecting other factors representing traffic environment, safety, and traffic efficiency may be used.

The CPU 31 may determine the state vector S that minimizes the objective function C(S) by substituting all possible state vectors S into the objective function C(S). A technique may be used to determine the state vector S that minimizes the objective function C(S). Furthermore, if C _multi (S) has less influence on the objective function C(S) than C _single (S) and C _consistency (S) on the objective function C(S), then the CPU 31 determines that the objective function A state vector S that minimizes the objective function C(S) may be determined by dividing C(S) into subproblems in which C _multi (S) is deleted. Also, the state vector _S is divided into S=(S _a , S _b ), where S _a =(s ₁ , . . . , s _i ) and S _b =(s _i+1 , . Then, if C _multi (S)=C _multi (S _a )+C _multi (S _b ), the CPU 31 assigns the objective function C(S) to the subproblems C(S _a ) and C(S _b ). By dividing, the state vectors S _a and S _b that minimize each may be determined.

In step S70 of FIG. 14, the CPU 31 generates a virtual traffic rule for the vehicle Q, which is an automatic vehicle, and the vehicle P, which is a manual vehicle, is the vehicle 7 that requires the virtual traffic rule, that is, the vehicle P is a quasi-compliant vehicle 7B. If so, a virtual traffic rule generation process for generating a virtual traffic rule for the vehicle P is executed. The virtual traffic rule generation processing will be described later in detail.

In step S80, the CPU 31 determines whether or not to generate control information and the content of the control information according to the types of the vehicle P and the vehicle Q. FIG.

Specifically, the CPU 31 generates control information including a global route and virtual traffic rules for the vehicle Q, which is an automobile. When the vehicle P is a manual vehicle requiring virtual traffic rules, that is, a semi-observant vehicle 7B, the CPU 31 generates control information for the vehicle P including the virtual traffic rules. If the vehicle P is a manual vehicle that does not require virtual traffic rules, that is, a non-observing vehicle 7C or a non-connected vehicle 7D, the CPU 31 does not generate control information for the vehicle P.

The CPU 31 performs control for transmitting the generated control information to each of the vehicles 7 for which the control information is generated, through the wireless communication device 3 .

In this case, the CPU 31 grasps the position of the vehicle 7 for which the control information is generated from the latest mobile object information, and transmits the control information from the wireless communication device 3 closest to the current position of the vehicle 7. , specifies the wireless communication device 3 that transmits the control information. The location information of the wireless communication device 3 is pre-stored in the non-volatile memory 34, for example.

Thus, the control processing shown in FIG. 14 ends.

Vehicles 7 that have received the control information drive in accordance with the virtual traffic rules included in the control information, so interference between vehicles 7 at the point of interference X is avoided.

Next, the virtual traffic rule generation processing in step S70 of FIG. 14, which was briefly mentioned earlier, will be described. FIG. 19 is a flowchart showing an example of the flow of virtual traffic rule generation processing.

In step S100, the CPU 31 determines each interference point X between the vehicle P and the vehicle Q specified in step S40 of FIG. 14, the interference time t at each interference point X calculated in step S50 of FIG. A virtual traffic rule is generated using the state s _x at each interference point X that has been set. Specifically, the CPU 31 determines the position information (for example, the interference point ID) indicating the position of each interference point X, the state s _{x at each interference point X, and each interference point X obtained from the state s x and} the interference time t. Generate a virtual traffic rule that associates the entry prohibition time with the time. The CPU 31 may include the range of the interference point X in the position information representing the position of each interference point X. FIG. The range of the interference point X is also referred to as a "no-entry range 20" because it is a range into which the vehicle 7 cannot enter during the no-entry time.

If the vehicle P and the vehicle Q enter the interfering point X according to the generated virtual traffic rule, the vehicle P and the vehicle Q can be prevented from interfering at the interfering point X. However, since the vehicle P is a manually operated vehicle, even if the virtual traffic rules for the vehicle P are generated, the driver of the vehicle P does not necessarily drive the vehicle P according to the virtual traffic rules. Therefore, it is assumed that the speed of the vehicle ^P suddenly changes and the collision time tp of the vehicle ^P at the collision point X deviates from the collision time tp calculated in step S50 of FIG.

FIG. 20 is a diagram illustrating an example of traffic flow caused by such a change in speed of the vehicle P. In FIG. In FIG. 20, the vehicle P is traveling on the priority road and the vehicle Q is traveling on the non-priority road.

Vehicle Q, which travels according to the virtual traffic rule in which the passing priority of vehicle P is set to priority at interference point X, stops before interference point X and waits until vehicle P passes interference point X (FIG. 20). (A)).

In this situation, for example, the vehicle P suddenly decelerates temporarily in an attempt to avoid an obstacle, and the speed of the vehicle P at time T represented by the mobile body information of the vehicle P becomes time T− 1, the time t ^p _begin at which the vehicle P begins to pass through the interference point X is delayed, and even though the vehicle P does not pass through the interference point X, the vehicle Q interferes A situation occurs in which the vehicle continues to stop before the point X.

Therefore, it is assumed that the control device 10 generates a new virtual traffic rule in which the passing priority of the vehicle Q is set to be higher than that of the vehicle Q, and transmits the new virtual traffic rule to the vehicle Q. FIG. Vehicle Q, which has received the new virtual traffic rule, starts traveling according to the virtual traffic rule and enters the interference point X (see FIG. 20(B)).

On the other hand, when the vehicle P, which has no trouble traveling by avoiding the obstacle, increases its speed, the time t ^p _begin at which the vehicle P begins to pass through the interference point X is hastened. For vehicle Q, a new virtual traffic rule is generated in which the passing priority of vehicle P is set to priority, and is transmitted to vehicle Q.

As a result, the vehicle Q may stop at the interfering point X to wait for the vehicle P to pass, thereby obstructing the passage of the vehicle P (see FIG. 20(C)).

FIG. 21 is a diagram showing an example of virtual traffic rules for the vehicle Q, which are generated with respect to the speed change of the vehicle P described in FIG.

FIG. 21A is a graph showing changes in speed of the vehicle P. Period IT ₁ represents the period before the vehicle P suddenly decelerates temporarily, and period IT ₂ represents the period during which the vehicle P suddenly decelerates. and a period _IT3 represents a period during which the vehicle P accelerates and increases its speed.

FIG. 21(B) is generated in step S100 of FIG. 19, in which the passing priority of vehicle Q is set higher _than that of vehicle Q _only during period IT2 as the speed of vehicle P changes during period IT2. represents an example of a virtual traffic rule. As described earlier _, when the vehicle Q enters the interference point X according to the virtual traffic rule shown in FIG.

Therefore, the control device 10 according to the present invention performs the processing after step S110 in _FIG . Generate enhanced virtual traffic rules based on a fail-safe design concept that considers the worst case of possible acceleration.

Specifically, the control device 10 predicts the time at which the vehicle P starts passing through the interference point X due to the deceleration of the vehicle P in the period IT2, like the speed change of the vehicle P shown in _FIG . An enhanced virtual traffic rule as shown in FIG. 21(C) is generated so as not to set the passage priority of vehicle Q to priority in period IT2 even if it is later _than time t ^p _begin .

Therefore, in step S110 of FIG. 19, the CPU 31 acquires the speed of the vehicle P from the moving body information of the vehicle P, which is a vehicle approaching the vehicle Q and is a manually operated vehicle. The speed of the vehicle P is an example of a state value related to the state of an approaching moving object approaching the vehicle Q. FIG. Since time has passed while the CPU 31 is acquiring the speed of the vehicle P, the moving body information of the vehicle P acquired in step S110 is already the moving body information of the vehicle P at the time T-1 one unit earlier. is represented.

In step S120, the CPU 31 predicts the possible speed range of the vehicle P at time T after one unit from the moving body information of the vehicle P at time T-1 one unit before obtained in step S110. set the predicted range of the speed of the vehicle P at . The upper limit speed of the set prediction range of speed is represented as ^vupper , and the lower limit speed is represented as ^vlower . For the speed prediction, the speed and acceleration of the vehicle P at time T−1 and the amount of operation of the accelerator and brake by the driver may be used to predict the speed of the vehicle P at time T. The speed of the vehicle P at time T may be predicted by performing machine learning using a support vector machine on learning data obtained by combining each past time and the speed of the vehicle P at that time.

In step S130, the CPU 31 acquires the current speed of the vehicle P from the moving body information of the vehicle P acquired at the timing corresponding to the time T.

In step S140, the CPU 31 determines whether or not the current speed of the vehicle P acquired in step S130 is within the predicted speed range set in step S120. If the current speed of the vehicle P is not within the prediction range, the process proceeds to step S150.

FIG. 22 is a diagram showing a state example in which the current speed of the vehicle P is not within the prediction range. The virtual traffic rule generated in step S100 of FIG. 19 is a virtual traffic rule generated on the premise that the current speed VT of the vehicle P at time _T falls within the prediction range. That is, as shown in FIG. 22, when the current speed V _T of the vehicle P at time T is not included in the prediction range predicted from the moving object information including the speed V _T-1 at time T-1, the When the vehicle Q travels according to the virtual traffic rule generated in step S100 of step _S100 , the vehicle Q interferes with the vehicle P at the interference point X, The probability of obstructing the traffic of the vehicle P at the interfering point X increases. Therefore, in step S150, the CPU 31 determines that the vehicle P has become abnormal.

Since there is an abnormality in vehicle P, CPU 31 generates a reinforced virtual traffic rule in step S160.

For example, the CPU 31 sets the entry prohibition time of the vehicle Q longer than the virtual traffic rule for the vehicle Q generated in step S100 in accordance with the speed change of the vehicle P, so that the vehicle Q interferes with the vehicle P at the interference point X. Generate enhanced virtual traffic rules with stronger constraints to make it harder.

FIG. 23 is a diagram showing an example of enhanced virtual traffic rules for vehicle Q. In FIG. As shown in FIG. 23, when the speed of the vehicle P is slower than the lower limit speed v ^lower of the prediction range, the CPU 31 extends the entry prohibition time of the vehicle Q with respect to the virtual traffic rule of the vehicle Q generated in step S100. Generate an enhanced virtual traffic rule extended backward by time δ. In FIG. 23, the original t ^p _end represents the t ^{p end of the interference time t p calculated before the vehicle P abnormality was detected, and the new t p} _end ^{represents the} _vehicle P abnormality detected. t ^p _end of the interference time t ^p assumed from the speed of the vehicle P at the time.

Further, when the speed of the vehicle P is faster than the upper limit speed v ^upper of the prediction range, the CPU 31 sets the entry prohibition time of the vehicle Q to the extension time δ for the virtual traffic rule of the vehicle Q generated in step S100 of FIG. Generates a reinforced virtual traffic rule that extends only forward. Naturally, the CPU 31 considers the possibility that the speed of the vehicle P will continue to change irregularly beyond the predicted range, and the virtual traffic rules for the vehicle Q generated in step S100 are applied to the virtual traffic rules for the vehicle Q. The entry prohibition time may be extended forward and backward by the extension time δ. In this case, the CPU 31 may set the extended time δ to be extended forward and the extended time δ to be extended backward to different values.

With the above, the virtual traffic rule generation processing shown in FIG. 19 ends.

On the other hand, when it is determined in the determination process of step S140 in FIG. 19 that the current speed _VT of the vehicle P is within the prediction range as shown in FIG. 24, the process proceeds to step S170.

In this case, if the vehicle Q travels according to the virtual traffic rules for the vehicle Q generated in step S100, the interference with the vehicle P at the interference point X can be avoided. Therefore, in step S170 of FIG. 19, the CPU 31 determines that the vehicle P is normal without any abnormality. Further, since there is no need to generate a reinforced virtual traffic rule because there is no abnormality in the vehicle P, the CPU 31 performs the virtual traffic rule generating process shown in FIG. 19 without generating a reinforced virtual traffic rule in step S160. finish.

In the virtual traffic rule generation process shown in FIG. 19, the speed of the vehicle P is used as the state value regarding the state of the vehicle P, which is the determination target for determining whether the vehicle P approaching the vehicle Q has an abnormality. However, instead of the speed of the vehicle P, for example, the acceleration of the vehicle P, the passage time t ^p during which the vehicle P passes through the interference point X, and the position of the vehicle P may be used. Furthermore, the CPU 31 sets prediction ranges for a plurality of state values, such as the speed of the vehicle P, the acceleration of the vehicle P, the passage time t ^p for the vehicle P to pass through the interference point X, and the position of the vehicle P, and When any one state value is not included in the prediction range, or when a predetermined number (for example, two) of state values are not included in the prediction range, it is determined that an abnormality has occurred in the vehicle P. may

For the prediction of each state value to be determined as an abnormality of the vehicle P, not only the moving body information of the vehicle P, but also traffic rule information included in the control travel route map 19, for example, restrictions on the road on which the vehicle P travels. Information representing the regulation and shape of the road on which the vehicle P is traveling, such as speed, number of lanes, curvature, width, slope, and lane priority, may be used.

In addition, when the vehicle P is a non-connected vehicle 7D, the abnormality of the vehicle P such as the speed of the vehicle P, the acceleration of the vehicle P, the passage time t ^p during which the vehicle P passes through the interference point X, and the position of the vehicle P The state value to be determined is measured based on the image captured by the collecting device 9 . Therefore, as the image quality of the image captured by the collecting device 9 deteriorates, the measurement accuracy of the state value to be determined as to whether the vehicle P is abnormal also deteriorates. If the presence or absence of an abnormality in the vehicle P is determined using a state value whose measurement accuracy has decreased, the reliability of the determination result will also decrease.

Therefore, even when it is determined in step S140 in FIG. If it is considered to be lower, it is preferable to determine that an abnormality has occurred in the vehicle P based on the fail-safe design concept.

Image quality is affected by weather conditions. For example, the image quality of an image captured by the collection device 9 in rain or fog may be lower than the image quality of an image captured by the collection device 9 in fine weather.

In addition, the image quality of the image is affected by the operating state of the collection device 9 . For example, the image quality of an image captured when the collecting device 9 has some kind of failure may be lower than the image quality of an image captured when there is no failure.

Therefore, when the amount of rainfall in the area where the collection device 9 is installed is greater than the reference value, when the occurrence of fog is notified in the area where the collection device 9 is installed, and when the failure notification is notified from the collection device 9 If at least one of the above cases occurs, the CPU 31 may determine that the measurement accuracy of the state value to be determined as an abnormality of the vehicle P is lower than the predetermined reference accuracy. The CPU 31 acquires weather information from an external device connected to the communication network 5, for example.

Further, when the error between the predicted state value and the actual state value is larger than a predetermined value, the CPU 31 determines that the measurement accuracy of the state value is lower than the predetermined reference accuracy. You may

Note that even if the vehicle P is the quasi-compliant vehicle 7B and the non-compliant vehicle 7C that do not measure the state value of the abnormality determination target of the vehicle P based on the image captured by the collection device 9, the predicted state value If the error of the actual state value is larger than the predetermined value, the CPU 31 may determine that the measurement accuracy of the state value is lower than the predetermined reference accuracy.

Until now, when the reinforced virtual traffic rule was generated in step S160 of FIG. An example of extending the prohibited time has been described. However, instead of extending the entry prohibition time, the CPU 31 sets the range of the interference point X, that is, the entry prohibition range 20 of the vehicle Q to the entry prohibition range 20 represented by the virtual traffic rule of the vehicle Q generated in step S100. It can be set wider than

FIG. 25 is a diagram showing an example of the no-entry area 20 set around the interference point X. In FIG. In FIG. 25, a no-entry range 20A represents the no-entry range 20 represented by the virtual traffic rules for the vehicle Q generated in step S100 of FIG. A forbidden range 20 is represented. In this way, by setting the entry prohibition range 20B wider than the entry prohibition range 20A, the vehicle P and the vehicle Q are less likely to interfere with each other at the interference point X.

Also, the CPU 31 may adjust the strength of the restriction in the enhanced virtual traffic rule according to the number of times an abnormality in the vehicle P is detected.

The fact that an abnormality is detected multiple times with respect to the same vehicle P means that the driver of the vehicle P tends to frequently change lanes 8 or repeat sudden stops and sudden starts. In other words, the more the number of times an abnormality is detected, the more difficult it is to predict the running state of the vehicle P. Therefore, even if the vehicle Q travels according to the virtual traffic rules, it will interfere with the vehicle P that has never detected an abnormality at the interference point X. higher probability.

Therefore, the CPU 31 sets the virtual traffic rules for the vehicle Q generated in step S100 of FIG. It is preferable to generate enhanced virtual traffic rules with increased constraints. For example, the CPU 31 may set the extension time .delta.

<Modified example of enhanced virtual traffic rule generation>
When the vehicle P, which is a manually operated vehicle, is a connected vehicle that travels according to the received virtual traffic rules, that is, when the vehicle P is the compliant vehicle 7A and the semi-compliant vehicle 7B, the CPU 31 controls not only the vehicle Q Enhanced virtual traffic rules may be generated for vehicle P as well.

By generating the reinforced virtual traffic rules for the vehicle P and the vehicle Q, it becomes more difficult for the vehicles P and Q to interfere at the interference point X than when the reinforced virtual traffic rule is generated only for the vehicle Q.

FIG. 26 is a diagram showing an example of a reinforced virtual traffic rule generated when an abnormality is detected in the vehicle P in a situation where the passing priority of the vehicle P is set to priority as shown in FIG. is.

As shown in FIG. 26, the CPU 31 generates an enhanced virtual traffic rule in which the entry prohibition time is set so as to include an overlap period in which the interference time tp and the interference time ^tq overlap with respect to the vehicle ^P and the vehicle Q. do. The no-entry time for vehicle P is denoted by "τ", where τ = t ^p _end - t ^q _begin + τ ₀ . Here, "τ ₀ " is a safety margin and takes a real value of 0 or more. As the safety margin τ ₀ increases, the time during which the vehicle P is prohibited from entering the interference point X increases, so the probability that the vehicle P and the vehicle Q interfere with each other at the interference point X can be further reduced.

Note that the safety margin τ ₀ is a value that is set in consideration of the shape of the road on which the vehicle P is traveling and the information representing the line-of-sight condition of the interference point X. For example, in the case of an interference point X with poor visibility, the safety margin τ ₀ is set larger than the reference value. In addition, the safety margin τ ₀ does not need to be a fixed value. ₀ can be even larger. Further, the CPU 31 may adjust the size of the safety margin τ ₀ according to the time zone in which the vehicle P travels, such as making the safety margin τ ₀ at night larger than the safety margin τ ₀ during the day.

On the other hand, for the vehicle Q, the CPU 31 generates an enhanced virtual traffic rule in which the entry prohibition time for the vehicle Q is extended by the extension time δ with respect to the virtual traffic rule for the vehicle Q generated in step S100 of FIG. The extension time .delta. in this case is expressed by .delta.=max(.tau., _.delta.min ). That is, the extension time δ is set to the longer one of the entry prohibition time τ for the vehicle P and the lower limit value δ _min of the extension time δ predetermined for the extension time δ.

It should be noted that the lower limit value δ _min of the extension time δ is also a value that is set in consideration of the shape of the road on which the vehicle Q is traveling and the information representing the line-of-sight situation, similarly to the safety margin τ ₀ . For example, in the case of the interference point X with poor visibility, it is preferable to set the lower limit δ _min larger than the reference value. Similarly to the safety margin τ ₀ , the lower limit value δ _min does not need to be a fixed value, and the lower limit value δ _min may be adjusted according to the weather conditions around the interference point X and the time period during which the vehicle Q travels.

As already explained, in the virtual traffic rules generated in step S100 of FIG. An exclusive setting is made such that the vehicle Q is set to be able to enter the interference point X over the entire interference time ^tp , and the vehicle Q is prohibited from entering the interference point X over the entire interference time ^tp .

On the other hand, when the modification related to the generation of this enhanced virtual traffic rule is applied, a time period is set during which both the vehicle ^P and the vehicle Q are prohibited from entering the interference point X within the interference time tp. . In this way, by generating the enhanced virtual traffic rules in which the entry prohibition times for the vehicle P and the vehicle Q are set to overlap, either the vehicle 7 of the vehicle P or the vehicle Q is allowed to comply with the enhanced virtual traffic rule. Even if the vehicles 7 erroneously enter the interfering point X without complying, the interference between the vehicles 7 at the interfering point X can be avoided.

Note that the farther the vehicle P is from the interference point X, the longer the vehicle P's interference time t ^p because the number of factors to be considered in calculating the interference time t ^p increases. Calculation accuracy of t ^p increases and becomes shorter. The same is true for vehicle Q's interference time t ^q . That is, the overlapping period in which the interference time t ^p and the interference time t ^q overlap becomes shorter as the vehicles P and Q approach the interference point X, respectively. If the overlapping period of the interference time t ^p and the interference time t ^q is shortened, the entry prohibition time τ for the vehicle P is shortened accordingly, and after the vehicle P traveling according to the enhanced virtual traffic rule stops before the interference point X, It will behave as if it will start immediately.

Such a short-term stop of about several seconds hinders the smooth running of the vehicle P, so the CPU 31 changes the content of the enhanced virtual traffic rule generation according to the length of the overlapping period of the interference time t ^p and the interference time t ^q . You may

Specifically, in a situation where the passing priority of the vehicle P at the interference point X is set to priority, when the entry prohibition time τ for the vehicle P exceeds the predetermined threshold _τth , as shown in FIG. , the CPU 31 generates an enhanced virtual traffic rule in which an entry prohibition time τ is set for the vehicle P, and an enhanced virtual traffic rule for the vehicle Q in which the entry prohibition time for the vehicle Q is extended by the extension time δ. .

On the other hand, in a situation where the priority order of passage of the vehicle P at the interference point X is set to priority, if the entry prohibition time τ for the vehicle P is equal to or less than the threshold _τth , the CPU 31 applies the enhanced virtual traffic rules for the vehicle P is not generated, and only for the vehicle Q, a reinforced virtual traffic rule is generated in which the entry prohibition time of the vehicle Q is extended by the extension time δ. That is, the CPU 31 does not set the entry prohibition time τ for the vehicle P, but extends the entry prohibition time for the vehicle Q by the extension time δ.

It should be noted that the vehicle P is approaching the vehicle Q when the vehicle P is approaching the vehicle Q to the point where the entry prohibition time τ is equal to or less than the threshold _τth .

FIG. 27 is a diagram showing an example of constraint conditions that constrain the running of the vehicle P and the vehicle Q when the entry prohibition time τ is equal to or less than the threshold _τth . The virtual traffic rule for vehicle P generated in step S100 of FIG. 19 is applied as the constraint condition for vehicle P, and the virtual traffic rule for vehicle Q generated in step S100 in FIG. A reinforced virtual traffic rule is applied in which Q's entry prohibition time is extended by the extension time δ.

In the modification related to the generation of the enhanced virtual traffic rule, the CPU 31 also sets the lower limit value δ _min of the extension time δ and the safety margin τ ₀ so that the extension time δ increases as the number of abnormalities detected in the vehicle P increases. At least one value of may be adjusted.

So far, the modified example related to the generation of the reinforced virtual traffic rule in the situation where the priority order of passage of the vehicle P at the interference point X has been set to priority has been described. , the CPU 31 sets the entry prohibition time so as to include the overlap period in which the interference time t ^p and the interference time t ^q overlap with respect to the vehicle P and the vehicle Q, and sets the enhanced virtual traffic rule to generate

FIG. 28 is a diagram showing an example of a reinforced virtual traffic rule generated when an abnormality is detected in the vehicle P in a situation where the passing priority of the vehicle P is set to non-prioritized.

In this case, the CPU 31 generates a reinforced virtual traffic rule including an entry prohibition time τ which is set so as to include an overlap period in which the interference time t ^p and the interference time t ^q overlap with respect to the vehicle Q. On the other hand, for the vehicle P, the CPU 31 generates an enhanced virtual traffic rule in which the entry prohibition time for the vehicle P is extended by the extension time δ with respect to the virtual traffic rule for the vehicle P generated in step S100 of FIG.

As a matter of course, when the entry prohibition time τ for the vehicle Q is equal to or less than the threshold τ _th , the CPU 31 does not generate an enhanced virtual traffic rule for the vehicle Q as shown in FIG. Only for this, a reinforced virtual traffic rule is generated in which the entry prohibition time of the vehicle P is extended by the extension time δ.

In the present embodiment, an example in which the control system 1 controls the vehicle 7 has been described. Mobile objects include pedestrians, bicycles, motorcycles, wheelchairs, kickboards, ships, and flying objects such as drones.

Similar to the vehicle 7, these mobile bodies are classified into compliant mobile bodies, semi-compliant mobile bodies, non-compliant mobile bodies, and non-connected mobile bodies according to the functions and states of the mobile bodies.

Although the present invention has been described using the embodiments, the present invention is not limited to the scope described in the embodiments. Various changes or improvements can be made to the embodiments without departing from the gist of the present invention, and forms with such changes or improvements are also included in the technical scope of the present invention. For example, the order of processing may be changed without departing from the gist of the present invention.

In the embodiment, as an example, the form in which the control processing in the control device 10 is implemented by software has been described. Programmable Gate Array) or PLD (Programmable Logic Device), and hardware processing may be performed. In this case, the speed of processing can be increased compared to the case where each processing is realized by software.

In this way, the CPU 31 of the control device 10 and the CPU 41 of the control device of the vehicle 7 are, for example, ASICs, FPGAs, PLDs, GPUs (Graphics Processing Units), and FPUs (Floating Point Units). can be replaced with

Further, the operations of the CPU 31 of the control device 10 and the CPU 41 of the control device of the vehicle 7 in the embodiment may be realized by a plurality of CPUs 31 and a plurality of CPUs 41 in addition to the form realized by one CPU 31 and one CPU 41, respectively. Furthermore, the operations of the CPU 31 of the control device 10 and the CPU 41 of the control device of the vehicle 7 in the embodiment are realized by the cooperation of the CPU 31 of the computer 30 and the CPU 41 of the computer 40, which are physically separated from each other. There may be.

In the above-described embodiment, the control program read by the CPU 31 of the control device 10 is installed in the ROM 32, and the control program read by the CPU 41 in the control device of the vehicle 7 is installed in the ROM 42. However, the configuration is limited to this. not to be The control program and control program according to the present invention can also be provided in a form recorded on a storage medium readable by the computer 30 and the computer 40 . For example, the control program and the control program may be provided in a form recorded on an optical disc such as CD (Compact Disc)-ROM or DVD (Digital Versatile Disc)-ROM. Also, the control program and the control program may be provided in a form recorded in a portable semiconductor memory such as a USB (Universal Serial Bus) memory or a memory card. ROM 32, ROM 42, nonvolatile memory 34, nonvolatile memory 44, CD-ROM, DVD-ROM, USB, and memory cards are examples of non-transitory storage media.

Further, the control device 10 and the control device of the vehicle 7 may download control programs and control programs from external devices connected to the communication network 5, respectively. In this case, the CPU 31 of the control device 10 and the CPU 41 of the control device of the vehicle 7 respectively read the control program and the control program downloaded from the external device and execute control processing and control processing of the vehicle 7 .

The following additional remarks are disclosed regarding the above embodiments.

(Appendix 1)
memory;
at least one processor connected to the memory;
including
The processor
receiving mobile information representing the state of each mobile, including autonomous mobiles that receive constraints defining travel time and space;
Among the moving bodies that have received the moving body information, an abnormality related to the movement of the approaching moving body is detected from a change in the state of the approaching moving body that may approach the autonomous moving body within a predetermined range in the interference area. detect whether or not it has occurred,
When an anomaly is detected, the constraint condition is stronger than the constraint condition generated before the anomaly detection unit detects an anomaly so that the approaching moving body is less likely to interfere with the interference area. A control device that generates a strengthening constraint condition for the autonomous mobile body.

(Appendix 2)
A non-temporary storage medium storing a program executable by a computer to execute virtual traffic rule generation processing,
The virtual traffic rule generation process includes:
Among mobile bodies that have received mobile body information representing the state of the mobile body, the autonomous mobile bodies in the interference area among the mobile bodies that include the autonomous mobile body that receives the constraint conditions that define the travel time and space Detecting whether or not an abnormality related to the movement of the approaching moving body has occurred from a change in the state of the approaching moving body that may approach within a predetermined range,
When an anomaly is detected, the constraint condition is stronger than the constraint condition generated before the anomaly detection unit detects an anomaly so that the approaching moving body is less likely to interfere with the interference area. A non-transitory storage medium that generates a hardening constraint for the autonomous mobile body.

1 control system, 3 wireless communication device, 5 communication network, 6 (6P, 6Q) route, 7 vehicle, 7A compliant vehicle, 7B semi-compliant vehicle, 7C non-compliant vehicle, 7D non-connected vehicle, 8 lane, 8A priority lane, 8B non-priority lane, 9 collection device, 9A infrastructure sensor, 9B state detection unit, 10 control device, 11 destination setting unit, 12 communication unit, 13 global route planning unit, 14 travel route prediction unit, 15 interference point identification unit, 16 anomaly detection unit 18 virtual traffic rule generation unit 19 control driving route map 20 (20A, 20B) no-entry area 21A (21B, 21C) position estimation unit 22A (22B, 22C) state management unit 23A ( 23B, 23C) wireless communication unit, 24 local route planning unit, 25A (25B, 25C) control unit, 26 vehicle travel route map, 27 virtual traffic rule transmission unit, 30 (40) computer, 31 (41) CPU, 32 ( 42) ROM, 33 (43) RAM, 34 (44) Nonvolatile memory, 35 (45) I/O, 36 (46) Bus, 37 (47) Communication unit, 38 (48) Input unit, 39 (49) Display unit, 51 internal sensor, 52 external sensor, 53 traveling device, δ extension time, lower limit of δ _min extension time, τ entry prohibition time, τ ₀ safety margin, _τth entry prohibition time threshold, C(S) Objective function, K1 to K7 points, S state vector, X interference point

Claims (11)

a receiving unit that receives moving body information representing the state of each moving body including an autonomous moving body that receives constraints defining travel time and space;
Among the moving bodies for which the receiving unit has received the moving body information, the approaching moving body can be determined from the change in the state of the approaching moving body, which has the possibility of approaching the autonomous moving body within a predetermined range in the interference area. an anomaly detection unit that detects whether an anomaly related to movement has occurred;
When an abnormality is detected by the abnormality detection unit, the constraint is strengthened so that the approaching moving body is less likely to interfere with the interference area than the constraint condition generated before the abnormality is detected by the abnormality detection unit. a generation unit that generates, for the autonomous mobile body, an enhanced constraint, which is the constraint, for the autonomous mobile body;
A control device with The abnormality detection unit detects that the state value of the state of the approaching moving object is not within a prediction range set using the moving body information of the approaching moving object received by the receiving unit. The control device according to claim 1, which detects that an abnormality related to movement has occurred. The abnormality detection unit detects at least one of the speed of the approaching moving body, the acceleration of the approaching moving body, the passage time for the approaching moving body to pass through the interference area, and the position of the approaching moving body. 3. The control device according to claim 2, wherein when the state value is not included in the prediction range set for the state, it is detected that an abnormality related to movement of the approaching moving object has occurred. The abnormality detection unit detects at least one of the speed of the approaching moving body, the acceleration of the approaching moving body, the passage time for the approaching moving body to pass through the interference area, and the position of the approaching moving body. When the measurement accuracy of the state value regarding the state is lower than a predetermined reference accuracy, it is detected that an abnormality related to movement of the approaching moving body has occurred even when the state value is within the prediction range. Item 2. The control device according to claim 2. The generating unit sets the enhanced constraint so that an entry prohibition time for prohibiting entry into the interference area is longer than the constraint generated before the abnormality is detected by the abnormality detection unit. The control device according to any one of claims 1 to 4, which is generated for an autonomous mobile object. The generation unit is configured to extend the time period during which entry into the interference area is prohibited under the strengthened constraint condition with respect to the time period during which entry into the interference area is prohibited under the constraint conditions generated before the abnormality is detected by the abnormality detection unit. The control device according to claim 5, wherein the extension time is adjusted using the shape of the interference area. When the approaching moving body is a connected moving body that receives the constraint conditions generated by the generating unit and moves according to the received constraint conditions,
The generating unit sets the strengthened constraint condition, which is set so that the autonomous mobile body and the connected mobile body are prohibited from entering the interference area for the same time, to the autonomous mobile body and the connected mobile body, respectively. 7. The control device according to claim 5 or 6, wherein the control device generates. As the number of detections of anomalies related to the movement of the approaching moving body increases, the generation unit increases the time period during which entry into the interference area is prohibited from the constraint conditions generated before the anomaly was detected by the anomaly detection unit. 8. The control device according to any one of claims 5 to 7, wherein the long set strong constraint condition is generated for the autonomous mobile body. The generator generates the strengthened constraint set such that the range prohibiting the autonomous mobile body from entering is wider than the constraint generated before the abnormality is detected by the abnormality detector. The control device according to any one of claims 1 to 8, which is generated for a body. The generating unit prohibits the autonomous moving object from entering under the constraint conditions generated before the abnormality is detected by the abnormality detecting unit as the number of detections of an abnormality related to the movement of the approaching moving object increases. The control device according to claim 9, wherein the strengthened constraint condition with a wide range of the interference area is generated for the autonomous mobile body. In an interference area, mobile body information representing the state of an approaching mobile body that may approach within a predetermined range from an autonomous mobile body that receives constraints defining travel time and space, and the state of the autonomous mobile body an acquisition device that acquires moving body information representing
A receiving unit for receiving moving body information of the autonomous moving body and the approaching moving body acquired by the acquiring device from the acquiring device, and a change in the state of the approaching moving body represented by the moving body information received by the receiving unit. an anomaly detection unit that detects whether an anomaly related to the movement of the approaching moving object has occurred, and if an anomaly is detected by the anomaly detection unit, the anomaly is generated before the anomaly is detected by the anomaly detection unit A generation unit that generates, for the autonomous mobile body, a strengthened constraint condition that is the constraint condition that strengthens the constraint so that the approaching moving body is less likely to interfere with the interference area than the constraint condition;
control system with

Images