Information Processing Method, Information Processing System, and Control Device

Abstract

An information processing method is an information processing method executed by a computer. In this information processing method, first conveyance object information is obtained. The first conveyance object information is conveyance object information including the layout and weight of the conveyance object in the moving body (S23). The weight of the moving body is obtained (S24). The obtained first conveyance object information and the weight of the moving body are input into a model (14d) generated using second conveyance object information, the weight of the moving body, and the rudder angle center value of the moving body, thereby generating information representing the predicted rudder angle center value. The second conveyance object information is past conveyance object information with respect to the first conveyance object information (S25). The information representing the predicted rudder angle center value is output to the moving body (S27).

Description

Technical Field

[0001] The present disclosure relates to an information processing method, an information processing system, and a control device related to a moving body. Background Art

[0002] In recent years, various studies have been conducted on autonomous vehicles. For example, a device has been developed to estimate the steering angle of an autonomous vehicle in order to perform driving control corresponding to the position of the center of gravity of the autonomous vehicle. A device that estimates and corrects errors of a plurality of sensors including a steering angle sensor is disclosed in Patent Document 1. In addition, a steering angle detection device is disclosed in Patent Document 2, which estimates an offset amount from the steering angle neutral point, that is, a zero point correction value, based on the appearance frequency of the steering angle during driving, and corrects the steering angle value using the zero point correction value.

[0003] (Prior Art Documents)

[0004] (Patent Documents)

[0005] Patent Document 1: Japanese Patent No. 6383907

[0006] Patent Document 2: Japanese Unexamined Patent Application Publication No. 2017-105294 Summary of the Invention

[0007] Problems to be Solved by the Invention

[0008] However, in the devices of Patent Document 1 and Patent Document 2, it takes time to set a reference value (hereinafter referred to as the steering angle center) of the steering angle of the moving body. For example, in Patent Document 1, the output values of a plurality of sensors including a steering angle sensor mounted on a vehicle are used as inputs, and the errors of the output values are estimated. In addition, in Patent Document 2, the steering angles detected during driving are accumulated, and the mode or average value is calculated as the steering angle center based on the accumulated plurality of steering angles. Thus, it takes time until convergence when estimating the steering angle center.

[0009] Therefore, an object of the present disclosure is to provide an information processing method, an information processing system, and a control device that can reduce the time required to set the steering angle center of a moving body.

[0010] Means for Solving the Problems

[0011] One aspect of the present disclosure relates to an information processing method executed by a computer. In the information processing method, first conveyance object information is obtained. The first conveyance object information is conveyance object information including the layout and weight of the conveyance object in the moving body. The weight of the moving body is obtained. The obtained first conveyance object information and the weight of the moving body are input into a model generated using second conveyance object information, the weight of the moving body, and the rudder angle center value of the moving body, thereby generating information representing a predicted rudder angle center value. The second conveyance object information is the conveyance object information in the past with respect to the first conveyance object information. The information representing the predicted rudder angle center value is output to the moving body.

[0012] Furthermore, one aspect of the present disclosure relates to an information processing system including: a first acquisition unit that acquires first conveyance object information, the first conveyance object information being conveyance object information including the layout and weight of the conveyance object in the moving body; a second acquisition unit that acquires the weight of the moving body; a generation unit that inputs the obtained first conveyance object information and the weight of the moving body into a model generated using second conveyance object information, the weight of the moving body, and the rudder angle center value of the moving body, thereby generating information representing a predicted rudder angle center value, the second conveyance object information being the conveyance object information in the past with respect to the first conveyance object information; and an output unit that outputs the information representing the predicted rudder angle center value to the moving body.

[0013] Furthermore, one aspect of the present disclosure relates to a control device mounted on a moving body. The control device includes: a first acquisition unit that acquires first conveyance object information, the first conveyance object information being conveyance object information including the layout and weight of the conveyance object in the moving body; a second acquisition unit that acquires the weight of the moving body; a generation unit that inputs the obtained first conveyance object information and the weight of the moving body into a model generated using second conveyance object information, the weight of the moving body, and the rudder angle center value of the moving body, thereby generating information representing a predicted rudder angle center value, the second conveyance object information being the conveyance object information in the past with respect to the first conveyance object information; and a setting unit that sets the predicted rudder angle center value as the rudder angle center value of the moving body.

[0014] Advantageous Effects of the Invention

[0015] By the information processing method and the like according to one aspect of the present disclosure, the time required to set the rudder angle center of the moving body can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a block diagram showing the functional configuration of the information processing system according to the embodiment.

[0017] Figure 2This is a diagram showing an example of reservation information of an embodiment.

[0018] Figure 3 This is a diagram showing an example of vehicle information of an embodiment.

[0019] Figure 4 This is a diagram showing an example of driving information of an embodiment.

[0020] Figure 5 This is a flowchart showing the operation of the information processing system of an embodiment.

[0021] Figure 6 This is a flowchart showing the details of the modeling process of an embodiment.

[0022] Figure 7A This is the first diagram for explaining the calculation of the offset of the center of gravity position of the vehicle.

[0023] Figure 7B This is the second diagram for explaining the calculation of the offset of the center of gravity position of the vehicle.

[0024] Figure 7C This is the third diagram for explaining the calculation of the offset of the center of gravity position of the vehicle.

[0025] Figure 8 This is a flowchart showing the details of the prediction process of an embodiment.

[0026] Figure 9 This is a flowchart showing the details of the dynamic estimation process of an embodiment.

[0027] Figure 10 This is a diagram for explaining the operation of the Kalman filter of an embodiment.

[0028] Figure 11 This is a block diagram showing the functional structure of the information processing system of Modification Example 1 of an embodiment.

[0029] Figure 12 This is a flowchart showing the operation of the information processing system of Modification Example 1 of an embodiment.

[0030] Figure 13 This is a block diagram showing the functional structure of the information processing system of Modification Example 2 of an embodiment.

[0031] Figure 14 This is a diagram showing an example of user information of Modification Example 2 of an embodiment.

[0032] Figure 15 This is a diagram showing an example of environmental information of Modification Example 2 of an embodiment.

[0033] Figure 16It is a flowchart showing the operation of the information processing system according to Modification Example 2 of the embodiment.

[0034] Figure 17 It is a block diagram showing the functional configuration of the information processing system according to Modification Example 3 of the embodiment.

[0035] Figure 18 It is a flowchart showing the determination process of Modification Example 3 of the embodiment.

[0036] Figure 19 It is a flowchart showing the details of the dynamic estimation process of Modification Example 3 of the embodiment.

[0037] Figure 20 It is a block diagram showing the functional configuration of the information processing system according to Modification Example 4 of the embodiment.

[0038] Figure 21 It is a flowchart showing the operation of the information processing system according to Modification Example 4 of the embodiment.

[0039] Figure 22 It is a block diagram showing the functional configuration of the information processing system according to Modification Example 5 of the embodiment.

[0040] Figure 23 It is a flowchart showing the operation of the information processing system according to Modification Example 5 of the embodiment.

[0041] Figure 24 It is a block diagram showing the functional configuration of the information processing system according to another embodiment. Detailed Embodiment

[0042] (Insight underlying the present disclosure)

[0043] An information processing method according to one aspect of the present disclosure is executed by a computer. In the information processing method, first conveyance object information is obtained, the first conveyance object information being conveyance object information including the layout and weight of the conveyance object in the moving body. The weight of the moving body is obtained. The obtained first conveyance object information and the weight of the moving body are input into a model generated using second conveyance object information, the weight of the moving body, and the rudder angle center value of the moving body, thereby generating information representing a predicted rudder angle center value. The second conveyance object information is past conveyance object information with respect to the first conveyance object information. The information representing the predicted rudder angle center value is output to the moving body.

[0044] Thus, by using a model generated using past conveyance information or the like, a predicted rudder angle center value can be generated. That is, without the moving body accumulating sensing data such as the rudder angle detected during travel, the rudder angle center value can be set. Thus, by means of the information processing method, the time required to set the rudder angle center of the moving body can be reduced. In addition, since the predicted rudder angle center value is generated using past conveyance information, the estimated value can be close to the rudder angle center value of the moving body in a state where the conveyance is actually loaded. Therefore, both the reduction in the correctness of the predicted rudder angle center value (in other words, maintaining the correctness) and the reduction in the time required to set the rudder angle center of the moving body can be achieved.

[0045] In addition, for example, the predicted rudder angle center value may be set as the initial value of the rudder angle center when the moving body travels corresponding to the first conveyance information.

[0046] Thus, a predicted rudder angle center value can be set before the travel of the moving body. The moving body controls the rudder angle according to the predicted rudder angle center value, so that the rudder angle can be correctly controlled even immediately after starting.

[0047] In addition, for example, in the conveyance information, at least the layout of the conveyance is the reservation information of the moving body, and the information indicating the predicted rudder angle center value is generated after the reservation information is determined. The reservation information may be the reservation information of the passengers of the moving body or the reservation information of the goods carried by the moving body.

[0048] Thus, the predicted rudder angle center value can be generated using the determined reservation information of the moving body. Since the reservation information is determined before travel, the predicted rudder angle center value can be generated before travel.

[0049] In addition, for example, in the conveyance information, at least one of the layout and weight of the conveyance is generated using the output data of the weight sensor provided in the moving body.

[0050] Thus, the layout or weight of the conveyance actually loaded on the moving body can be obtained based on the output data of the weight sensor, so that the predicted rudder angle center value of the moving body can be generated more correctly.

[0051] In addition, for example, in the conveyance information, at least one of the layout and weight of the conveyance is generated using the detection result of the conveyance, and the detection result of the conveyance is a detection result based on the output data of the imaging sensor provided in the moving body.

[0052] Thus, the layout or weight of the conveyance actually loaded on the moving body can be obtained based on the captured image, so that the predicted rudder angle center value of the moving body can be generated more correctly.

[0053] Alternatively, for example, the detection result may include the identification result of the transported object, and in the transported object information, at least one of the layout and weight of the transported object is generated using the attribute information of the identified transported object.

[0054] Thus, the layout or weight of the transported object can be generated using the attribute information of each transported object, so that the predicted rudder angle center value can be generated more accurately.

[0055] Alternatively, for example, the model is a model generated using a predefined basic model.

[0056] Thus, since the basic model is predetermined, the model can be generated by inputting specified information into the basic model. In other words, the model can be generated relatively easily.

[0057] Alternatively, for example, the model is a model generated using the center of gravity position of the moving body, the rudder angle center value, and the basic model, and the center of gravity position of the moving body is calculated based on the second transported object information and the weight of the moving body.

[0058] Thus, the model can be generated using the center of gravity position of the moving body and the rudder angle center value that can be calculated based on the transported object information and the weight of the moving body.

[0059] Alternatively, for example, the model is a model generated by performing machine learning using the second transported object information and the weight of the moving body as training data and the rudder angle center value as correct answer data.

[0060] Thus, a model with higher prediction accuracy than a pre-designed mathematical model can be generated.

[0061] Alternatively, for example, the model is a model generated using the training data further including the attribute information of the transported object.

[0062] Thus, a model considering the attribute information of the transported object can be generated, so that a model with further improved prediction accuracy can be generated.

[0063] Alternatively, for example, the model is a model generated using the training data further including the external environment information of the moving body.

[0064] Thus, a model considering the external environment information of the moving body can be generated, so that a model with further improved prediction accuracy can be generated.

[0065] In addition, an information processing system according to an aspect of the present disclosure includes: a first acquisition unit that acquires first conveyance object information, which is conveyance object information including the layout and weight of a conveyance object in a moving body; a second acquisition unit that acquires the weight of the moving body; a generation unit that inputs the acquired first conveyance object information and the weight of the moving body into a model generated using second conveyance object information, the weight of the moving body, and the center value of the rudder angle of the moving body, thereby generating information representing a predicted center value of the rudder angle, where the second conveyance object information is past conveyance object information with respect to the first conveyance object information; and an output unit that outputs the information representing the predicted center value of the rudder angle to the moving body. In addition, a control device according to an aspect of the present disclosure is mounted on a moving body, and the control device includes: a first acquisition unit that acquires first conveyance object information, which is conveyance object information including the layout and weight of a conveyance object in the moving body; a second acquisition unit that acquires the weight of the moving body; a generation unit that inputs the acquired first conveyance object information and the weight of the moving body into a model generated using second conveyance object information, the weight of the moving body, and the center value of the rudder angle of the moving body, thereby generating information representing a predicted center value of the rudder angle, where the second conveyance object information is past conveyance object information with respect to the first conveyance object information; and a setting unit that sets the predicted center value of the rudder angle as the center value of the rudder angle of the moving body.

[0066] Thus, it has the same effect as the above-described information processing method.

[0067] In addition, these general or specific aspects can be implemented by a system, a device, a method, an integrated circuit, a computer program, or a non-transitory recording medium such as a computer-readable CD-ROM, or can be implemented by arbitrarily combining a system, a device, a method, an integrated circuit, a computer program, and a recording medium.

[0068] Next, specific examples of an information processing method, an information processing system, and a control device according to an aspect of the present disclosure will be described with reference to the accompanying drawings. The embodiments shown here are all specific examples of the present disclosure. Thus, the numerical values, shapes, components, steps, order of steps, etc. shown in the following embodiments are all examples, and the gist thereof is not to limit the present disclosure. In addition, among the components of the following embodiments, those not described in the independent technical solutions are described as optional components. In addition, in all embodiments, the respective contents can be combined.

[0069] In addition, each drawing is a schematic diagram and not a precise illustration. Therefore, for example, the scales in each drawing are not the same. In addition, in each drawing, the same reference numerals are given to substantially the same components, and repeated descriptions are omitted or simplified.

[0070] In addition, in this specification, numerical values and numerical ranges not only represent strict expressions, but also represent ranges that are practically equivalent, for example, expressions including differences of about several percent.

[0071] (Embodiment)

[0072] Regarding the information processing method and the like related to this embodiment below, with reference to Figures 1 to 10 it will be described.

[0073] [1. Structure of Information Processing System]

[0074] First, regarding the structure of the information processing system 1 related to this embodiment, with reference to Figures 1 to 4 it will be described. Figure 1 is a block diagram showing the functional structure of the information processing system 1 of this embodiment.

[0075] As Figure 1 shown, the information processing system 1 includes a prediction device 10 and a vehicle 20. The prediction device 10 and the vehicle 20 are connected in a manner that enables communication via a network (not shown). The information processing system 1 is a prediction system that predicts the center value of the steering angle of the vehicle 20. The vehicle 20 transports passengers as the objects to be transported.

[0076] The prediction device 10 performs processing for predicting the center value of the steering angle of the vehicle 20. The prediction device 10, for example, uses the reservation information of the vehicle 20 (refer to Figure 2 described later) to predict the center value of the steering angle of the vehicle 20. The prediction device 10 is implemented by, for example, a server device.

[0077] The prediction device 10 includes a communication unit 11, a modeling unit 12, a first prediction unit 13, and an information management unit 14.

[0078] The communication unit 11 is used for the prediction device 10 to communicate with the vehicle 20. The communication unit 11 is implemented by, for example, a communication circuit (communication module). The communication unit 11 outputs the information representing the center value of the steering angle generated by the first prediction unit 13 to the vehicle 20 via a network (not shown), and obtains the detection results of the sensor unit 23 and the like from the vehicle 20. In this way, the communication unit 11 functions as an output unit that outputs the information representing the center value of the steering angle generated by the first prediction unit 13 to the vehicle 20. In addition, the communication unit 11 can obtain the driving information 14c from the vehicle 20. In addition, sometimes the center value of the steering angle predicted by the first prediction unit 13 is recorded as the predicted steering angle center value. In addition, sometimes generating the predicted steering angle center value (that is, predicting the center value of the steering angle) is recorded as predicting the predicted steering angle center value.

[0079] In addition, the communication unit 11 obtains various information such as reservation information 14a and vehicle information 14b of the vehicle 20 from an external device via a network. The various information obtained via the communication unit 11 is stored in the information management unit 14. The external device may be any device that is communicably connected to the prediction device 10 and is not particularly limited. For example, it may be a smartphone held by a passenger, or a smart speaker, a personal computer, or a cloud server. In this way, the communication unit 11 functions as a first acquisition unit for acquiring the reservation information 14a and also functions as a second acquisition unit for acquiring the vehicle information 14b. The reservation information 14a is an example of at least the layout of the passengers in the passenger information. In addition, the reservation information 14a may also include the weight of the passengers in the passenger information. In the present embodiment, the reservation information 14a includes the layout and weight of the passengers in the passenger information. In addition, the passenger information is an example of the freight information.

[0080] The modeling unit 12 generates a model 14d for the first prediction unit 13 to predict the predicted rudder angle center value of the vehicle 20 based on the various information stored in the information management unit 14. The modeling unit 12 generates the model 14d, for example, using a predefined basic model. In the present embodiment, the modeling unit 12 generates the model 14d using the past reservation information 14a, the weight of the vehicle 20, and the estimated value of the rudder angle center when the vehicle 20 travels based on the past reservation information 14a. The modeling unit 12 models the relationship between the center of gravity position calculated based on the past reservation information 14a and the vehicle 20 and the rudder angle center. In addition, the basic model is stored in the information management unit 14, for example.

[0081] In the present embodiment, the basic model is a linear regression model, and the modeling unit 12 generates the model 14d through regression analysis. The modeling unit 12 stores the generated model 14d in the information management unit 14. The modeling unit 12 generates the model 14d using the past reservation information 14a. Thus, the modeling unit 12 can generate the model 14d for predicting the predicted rudder angle center value of the vehicle 20 before the vehicle 20 travels.

[0082] In addition, the modeling unit 12 may generate the model 14d for each vehicle 20. The modeling unit 12 generates the model 14d, for example, for each vehicle ID that identifies the vehicle 20 (for example, refer to Figure 2 ).

[0083] In addition, the generation of the model 14d in the modeling unit 12 will be described later. Furthermore, the past reservation information 14a refers to the reservation information for the movement (travel) of the vehicle 20 based on the reservation content in the past. For example, a movement based on the reservation information 14a was made, and the reservation information 14a for the driving state during that movement was obtained, which is the past reservation information 14a. The driving state includes at least the estimated value of the steering angle center or information capable of calculating the estimated value of the steering angle center. The past reservation information 14a is an example of the second passenger information. The second passenger information is the past passenger information, such as the passenger information for which the vehicle 20 has already made a movement based on this passenger information at the current moment. In addition, the second passenger information is an example of the second transported object information.

[0084] The first prediction unit 13 uses the model 14d generated by the modeling unit 12 to predict the predicted steering angle center value of the vehicle 20. Specifically, the first prediction unit 13 inputs the current reservation information 14a obtained via the communication unit 11 and the weight of the vehicle 20 into the model 14d to generate information representing the predicted steering angle center value of the vehicle 20. The first prediction unit 13 predicts the predicted steering angle center value of the vehicle 20, for example, after the current reservation information 14a is determined. In other words, after the current reservation information 14a is determined, the predicted steering angle center value of the vehicle 20 is predicted. In addition, the first prediction unit 13 can predict the predicted steering angle center value of the vehicle 20, for example, after obtaining the current reservation information 14a and before the start of the travel based on this reservation information 14a. The start of the travel based on the reservation information 14a means that, as the start of the travel corresponding to the passenger information, for example, when the vehicle 20 starts to move in the state where the "time" included in the reservation information 14a is reached and the passenger is sitting in the "riding position". In addition, the first prediction unit 13 is an example of the generation unit that generates information representing the predicted steering angle center value. The current reservation information 14a is the reservation information 14a for which the vehicle 20 has not yet moved based on the reservation content, and is an example of the first passenger information. In addition, the first passenger information is an example of the first transported object information.

[0085] In addition, the first prediction unit 13 can, for example, predict the predicted steering angle center value before the vehicle 20 starts to move according to the reservation content, and output the information representing the predicted predicted steering angle center value to the vehicle 20. The first prediction unit 13 can use the model 14d generated by the modeling unit 12 before the travel of the vehicle 20 to predict the predicted steering angle center value, and thus output the information representing this predicted steering angle center value to the vehicle 20 before the vehicle 20 departs. In addition, the prediction made before the vehicle 20 departs is sometimes referred to as prior prediction.

[0086] In addition, the first prediction unit 13 is not limited to predicting the predicted rudder angle center value in advance. The first prediction unit 13 can predict the predicted rudder angle center value after the vehicle 20 departs and output it to the vehicle 20. In this case, the first prediction unit 13 can predict the predicted rudder angle center value immediately after the vehicle 20 departs and output it to the vehicle 20.

[0087] The first prediction unit 13 outputs the predicted predicted rudder angle center value to the vehicle 20 via the communication unit 11.

[0088] The information management unit 14 stores various information obtained via the communication unit 11, the model 14d used by the first prediction unit 13 to predict the predicted rudder angle center value, and the like. Specifically, the information management unit 14 stores reservation information 14a, vehicle information 14b, travel information 14c, and the model 14d.

[0089] The reservation information 14a includes the reservation details when the user reserves the vehicle 20. The reservation information 14a, for example Figure 2 as shown, includes "vehicle ID", "time", "departure location", "destination", and "seating position". Figure 2 FIG. is a diagram showing an example of the reservation information 14a of the present embodiment. In addition, Figure 2 represents the reservation information 14a in the case where the vehicle 20 is a four-seater vehicle that can accommodate four people.

[0090] The "vehicle ID" is identification information for determining the vehicle 20 reserved by the user. The "vehicle ID" can be a numerical value, a mark, or the like for determining the vehicle 20, or it can be a vehicle type name.

[0091] The "time" indicates the date or time when the vehicle 20 is used.

[0092] The "departure location" and "destination" are the locations where the passenger boards and alights from the vehicle 20.

[0093] The "seating position" indicates the layout of the passengers in the vehicle 20. Seats A to D represent the seats in the vehicle 20. Taking the case where the "vehicle ID" is 002 as an example, there are a total of 2 passengers. Specifically, the following is reserved: User 1 as a passenger sits in seat D, and User 2 as a passenger sits in seat B. For example, User 1 is a passenger sitting in seat D on the left side of the front row, and User 2 is a passenger sitting in seat B on the left side of the rear row. In addition, the "seating position" includes, for example, information for determining the weight of each user. The information for determining the weight can be the user's weight itself, or information such as the user's age and gender that can estimate the weight.

[0094] Thus, the reservation information 14a includes the layout and weight of the passengers in the vehicle 20. Additionally, in the case where the reservation information 14a does not include the weight of the passengers, a preset weight of the passengers can be used. Furthermore, the weight of the passengers can be calculated by using the above-described information for determining the weight, so as to be obtained as passenger information.

[0095] In addition, the information management unit 14 can store programs executed by each processing unit (e.g., the modeling unit 12 and the first prediction unit 13) included in the prediction device 10, as well as information for the execution of the programs. The information for executing the programs includes, for example, the basic model. The information management unit 14 is implemented by a storage device such as a semiconductor memory, for example.

[0096] The vehicle information 14b includes information for calculating the center of gravity position of the vehicle 20. The vehicle information 14b, for example Figure 3 as shown, includes "vehicle ID", "vehicle weight", "distance from the center of gravity of the vehicle weight to the front axle", "distance from the center of gravity of the vehicle weight to the rear axle", "distance from the center of gravity of the vehicle weight to the left wheel center", and "distance from the center of gravity of the vehicle weight to the right wheel center". Figure 3 is a diagram showing an example of the vehicle information 14b of the present embodiment.

[0097] The "vehicle ID" is identification information for determining the vehicle 20 reserved by the user. The "vehicle ID" in the vehicle information 14b is set to the same ID as the "vehicle ID" in the reservation information 14a for the same vehicle 20, for example. Thus, the modeling unit 12 can obtain information for generating the following model 14d from the vehicle information 14b, and the model 14d is a model suitable for the vehicle 20 reserved by the reservation information 14a.

[0098] The "vehicle weight" represents the weight of the vehicle 20 itself. In other words, the "vehicle weight" represents the weight of the vehicle 20 when no one is riding.

[0099] The "distance from the center of gravity of the vehicle weight to the front axle" represents the distance from the center of gravity of the vehicle weight of the vehicle 20 when no one is riding to the front axle. The "distance from the center of gravity of the vehicle weight to the rear axle" represents the distance from the center of gravity of the vehicle weight of the vehicle 20 when no one is riding to the rear axle. The "distance from the center of gravity of the vehicle weight to the left wheel center" represents the distance from the center of gravity of the vehicle weight of the vehicle 20 when no one is riding to the left wheel center. The "distance from the center of gravity of the vehicle weight to the right wheel center" represents the distance from the center of gravity of the vehicle weight of the vehicle 20 when no one is riding to the right wheel center. Additionally, the left wheel center is the center of the straight line connecting the tires arranged on the left side, specifically, the center of the straight line including the left front wheel and the left rear wheel. The right wheel center is the center of the straight line connecting the tires arranged on the right side, specifically, the center of the straight line including the right front wheel and the right rear wheel.

[0100] In addition, the vehicle information 14b may include information indicating the position of the center of gravity of the vehicle 20.

[0101] The vehicle information 14b may be obtained, for example, from an external device via the communication unit 11, or may be obtained from the vehicle 20 when the vehicle information 14b is stored in the vehicle 20.

[0102] The driving information 14c includes information indicating the driving history of the vehicle 20 when the vehicle 20 travels according to the reservation information 14a. The driving information 14c is generated, for example, based on the sensing results of various sensors mounted on the vehicle 20. The driving information 14c, for example Figure 4 as shown, includes "vehicle ID", "time", "vehicle speed", "steering angle", "estimated value of the steering angle center", and "straight driving judgment result". Figure 4 FIG. is an example showing the driving information 14c of the present embodiment.

[0103] The "vehicle ID" is identification information for identifying the vehicle 20 for which the sensing result has been obtained. The "vehicle ID" in the driving information 14c is set, for example, to the same ID as the "vehicle ID" in the reservation information 14a. Thus, the modeling unit 12 can obtain information for generating the following model 14d from the driving information 14c, and the model 14d is a model suitable for the vehicle 20 reserved by the current reservation information 14a.

[0104] The "time" indicates the time when the sensor unit 23 performs sensing during the driving of the vehicle 20. Here, an example is shown in which the sensor unit 23 performs sensing at intervals of 1 second, but it is not limited to this. In addition, an example is shown in which each of the various sensors included in the sensor unit 23 performs sensing at intervals of 1 second, but it is not limited to this.

[0105] The "vehicle speed" indicates the speed of the vehicle 20 at each moment. The "vehicle speed" is, for example, the sensing result of the vehicle speed sensor 23c.

[0106] The "steering angle" indicates the steering angle of the vehicle 20 at each moment. The "steering angle" is, for example, the sensing result of the steering angle sensor 23b.

[0107] The "estimated value of the steering angle center" indicates the estimated value of the steering angle center calculated for each moment of the vehicle 20. The "estimated value of the steering angle center" represents, for example, the estimation result of the dynamic estimation unit 22a.

[0108] The "straight driving judgment result" indicates the judgment result as to whether the vehicle 20 is traveling straight at each moment. For example, when the information processing system 1 includes a judgment unit ( Figure 1 not shown in the figure) for judging whether the vehicle is traveling straight, it includes the "straight driving judgment result". In addition, the "straight driving judgment result" may not be included in the driving information 14c.

[0109] The driving information 14c is obtained, for example, from the vehicle 20 via the communication unit 11. The driving information 14c can be obtained sequentially during the driving of the vehicle 20 based on the reservation information 14a, or can be obtained after the driving of the vehicle 20 based on the reservation information 14a. The driving information 14c is stored in correspondence with the reservation information 14a at the time when the driving information 14c is obtained. In other words, the information management unit 14 associates and stores the reservation information 14a with the driving information 14c obtained during the driving of the vehicle 20 based on the reservation information 14a.

[0110] Refer again to Figure 1 , the model 14d is a mathematical model generated by the modeling unit 12 for predicting the predicted steering angle center value of the vehicle 20. The model 14d is, for example, a model generated for each vehicle 20. In this case, the information management unit 14 can store the vehicle ID in a manner corresponding to the model 14d suitable for the vehicle 20 corresponding to the vehicle ID. Thus, the first prediction unit 13 can obtain the model 14d corresponding to the vehicle ID from the information management unit 14 based on the vehicle ID included in the reservation information 14a.

[0111] The vehicle 20 is an example of a moving body in which a user rides. The vehicle 20 is assumed to be an autonomous vehicle, that is, the driving of the vehicle is controlled without the operation of the driver. However, the vehicle 20 can also be a vehicle that can travel in either an autonomous driving or a manual driving mode, or can be a vehicle that can only be manually driven. In the case of a vehicle that can only be manually driven, the predicted steering angle center value is used for driving support processing to support manual driving.

[0112] The vehicle 20 includes a communication unit 21, a vehicle control unit 22, and a sensor unit 23.

[0113] The communication unit 21 is used for the vehicle 20 to communicate with the prediction device 10. The communication unit 11 is implemented, for example, by a communication circuit (communication module). The communication unit 21 obtains the predicted steering angle center value from the prediction device 10 via a network (not shown), and outputs the sensing result of the sensor unit 23 and the like to the prediction device 10.

[0114] The vehicle control unit 22 controls the driving of the vehicle 20. In the present embodiment, the vehicle control unit 22 controls the steering of the vehicle 20 based on the predicted steering angle center value obtained from the prediction device 10. The vehicle control unit 22 includes a dynamic estimation unit 22a and a steering control unit 22b. In addition, the vehicle control unit 22 controls acceleration, deceleration, and other controls in addition to steering.

[0115] The dynamic estimation unit 22a estimates the center value of the steering angle during the running of the vehicle 20. In other words, the dynamic estimation unit 22a generates an estimated value of the steering angle center. The dynamic estimation unit 22a sets the predicted steering angle center value obtained from the prediction device 10 as the steering angle center value of the vehicle 20. For example, the dynamic estimation unit 22a sets the predicted steering angle center value as the initial value of the steering angle center of the vehicle 20 before running to estimate the center value of the steering angle during the running of the vehicle 20. For example, the dynamic estimation unit 22a estimates the center value of the steering angle during the running of the vehicle 20 through Kalman filter processing. The steering angle center value can be expressed as a relative value such as a difference from a reference value or as an absolute value. The processing of the dynamic estimation unit 22a will be described later. In addition, the dynamic estimation unit 22a is an example of a setting unit.

[0116] The steering control unit 22b controls the steering of the vehicle 20 according to the estimation result of the dynamic estimation unit 22a. For example, the steering control unit 22b drives the steering electric motor according to the estimation result of the dynamic estimation unit 22a to change the angle of the wheels (for example, the front wheels) relative to the vehicle body. The steering control unit 22b further changes the angle of the wheels according to the depression degree of the accelerator, the depression degree of the brake, the vehicle speed, the angular velocity, etc.

[0117] In addition, the vehicle control unit 22 outputs the detection result obtained from the sensor unit 23 to the prediction device 10 via the communication unit 21.

[0118] The sensor unit 23 detects the running state of the vehicle 20. The sensor unit 23 has one or more sensors capable of detecting the state of the running vehicle 20. For example, it has a gyro sensor 23a, a steering angle sensor 23b, and a vehicle speed sensor 23c.

[0119] The gyro sensor 23a is a sensor that at least detects the angular velocity corresponding to the rotation around the vertical axis (yaw angle direction, for example, the Z-axis direction around described later Figure 7A ), that is, the angular velocity corresponding to the rotation (rotation) of the vehicle 20 in the left and right directions. The gyro sensor 23a can also be, for example, a sensor that detects the angular velocity of the rotation around each of the three axes with different directions of the vehicle 20 as axes, that is, the angular velocity of the rotation around each of the three axes including the vertical axis.

[0120] The steering angle sensor 23b is a sensor that detects the steering angle.

[0121] The vehicle speed sensor 23c is a sensor that detects the vehicle speed of the vehicle 20. The vehicle speed sensor 23c can be, for example, a sensor that detects the rotational speed of the axle per unit time and obtains the vehicle speed of the vehicle 20 according to the detected rotational speed.

[0122] The various sensors included in the sensor unit 23 output the detected detection results to the vehicle control unit 22. Additionally, the various sensors can perform detections synchronously or, for each sensor, at specified time intervals.

[0123] Furthermore, the sensor unit 23 may also include sensors other than those for detecting the driving state of the vehicle 20. For example, the sensor unit 23 may include a camera for photographing the interior and exterior of the vehicle 20, an acceleration sensor for detecting the acceleration of the vehicle 20, and a GPS (Global Positioning System) receiver for detecting the position of the vehicle 20.

[0124] As described above, the information processing system 1 according to the present embodiment includes: a modeling unit 12 that generates a model 14d for predicting the predicted steering angle center value of the vehicle 20 using past reservation information 14a, etc.; a first prediction unit 13 that predicts the predicted steering angle center value of the vehicle 20 using the model 14d generated by the modeling unit 12, the current reservation information 14a, and the vehicle information 14b; and a communication unit 11 that outputs the predicted predicted steering angle center value to the vehicle 20.

[0125] Accordingly, the information processing system 1 can generate the predicted steering angle center value using the model 14d generated by utilizing past reservation information 14a, etc. Moreover, by outputting the predicted steering angle center value to the vehicle 20 via the communication unit 11, the vehicle 20 can estimate the steering angle center value using the obtained predicted steering angle center value. For example, the vehicle 20 can estimate the steering angle center value using the predicted steering angle center value without accumulating the steering angles detected during driving. Therefore, the information processing method can shorten the time until the steering angle center of the vehicle 20 is estimated.

[0126] In addition, the dynamic estimation unit 22a of the vehicle 20 can, for example, set the predicted steering angle center value as the initial value of the steering angle center before the start of driving for the driving corresponding to the first passenger information performed by the vehicle itself.

[0127] Accordingly, since the dynamic estimation unit 22a can predict the steering angle center value before the start of driving of the vehicle itself, the running vehicle speed immediately after departure can be increased. For example, the dynamic estimation unit 22a can estimate the steering angle center in real time immediately after departure.

[0128] [2. Operation of the Information Processing System]

[0129] Next, the operation of the above-described information processing system 1 will be described with reference to Figures 5 to 10 as follows. Figure 5 is a flowchart showing the operation of the information processing system 1 of the present embodiment.

[0130] As Figure 5 shown, the information processing system 1 first performs a modeling process (S10). The modeling process is a process for generating a model 14d that predicts the predicted rudder angle center value of the vehicle 20. Further, when the reservation information 14a is determined, the information processing system 1 performs a prediction process (S20). The prediction process is a process that is executed after the current reservation information 14a is obtained, and uses the model 14d, the current reservation information 14a, and the vehicle information 14b to predict the predicted rudder angle center value of the vehicle 20. For example, the reservation process uses the model 14d, the user's seating position and weight, and the weight of the vehicle 20 to predict the predicted rudder angle center value of the vehicle 20. Further, after the information processing system 1 predicts the predicted rudder angle center value, when the vehicle 20 starts to travel, a dynamic estimation process (S30) is performed. The dynamic estimation process is a process of estimating the rudder angle center value of the traveling vehicle 20 using, for example, the predicted rudder angle center value predicted in step S20 as an initial value. The dynamic estimation process repeatedly estimates the rudder angle center value, for example, while the vehicle 20 is traveling.

[0131] In addition, the timing of performing the process of step S20 is not particularly limited as long as it is within the period from after the modeling process is performed to until the vehicle travels based on the reservation information 14a.

[0132] Next, the modeling process, the prediction process, and the dynamic estimation process will be described with reference to Figures 6 to 10 FIG. First, the modeling process will be described with reference to Figures 6 to 7C FIG. Figure 6 FIG. is a flowchart showing the details of the modeling process (S10) of the present embodiment. Figure 6 The process shown is executed by, for example, the modeling unit 12.

[0133] As Figure 6 shown, in the modeling process (S10), the modeling unit 12 reads various data from the information management unit 14 (S11). The modeling unit 12 reads, for example, past reservation information 14a, vehicle information 14b, and travel information 14c from the information management unit 14.

[0134] Next, the modeling unit 12 determines whether the number of samples of various data read out is greater than the first threshold (S12). In step S12, the modeling unit 12 determines whether the number of data it has is the number of samples sufficient to generate the model 14d. The first threshold is not particularly limited as long as it is the number of samples with which the modeling unit 12 can generate the model 14d. When the reservation information 14a and the driving information 14c that are associated with each other are regarded as one combination, the number of samples is defined as the number of combinations of the reservation information 14a and the driving information 14c for one vehicle ID. In addition, the number of samples may also be the cumulative time for obtaining the driving information 14c for one vehicle ID. When the cumulative time for obtaining the driving information 14c is longer than the specified time, the modeling unit 12 may determine that the number of samples is greater than the first threshold. By making the determination in step S12, the modeling unit 12 can generate the model 14d using the driving information 14c in various situations.

[0135] When the modeling unit 12 determines that the number of samples is greater than the first threshold ( "Yes" in S12), it calculates the center of gravity position (S13). The modeling unit 12 calculates the center of gravity position of the vehicle 20 when a passenger is in the vehicle (hereinafter referred to as the vehicle 20 during riding) using the past reservation information 14a and the vehicle information 14b. Specifically, the modeling unit 12 calculates the center of gravity position of the vehicle 20 during riding using the seating position and weight of the passenger and the center of gravity position and weight of the vehicle 20.

[0136] Here, regarding the calculation of the center of gravity position of the vehicle 20 during riding, reference Figures 7A to 7C will be made for explanation. Figure 7A FIG. 1 is a schematic view of the vehicle 20 when viewed from above, which is used to explain the calculation of the center of gravity position of the vehicle 20 during riding. Figure 7B FIG. 2 is a schematic view of the vehicle 20 when viewed from the rear, which is used to explain the calculation of the center of gravity position of the vehicle 20 during riding. Figure 7C FIG. 3 is a schematic view of the vehicle 20 when viewed from the side, which is used to explain the calculation of the center of gravity position of the vehicle 20 during riding.

[0137] In Figures 7A to 7C the front wheels 24a and 24b and the rear wheels 24c and 24d of the vehicle 20 are illustrated. The Y-axis is the front-rear direction of the vehicle 20, the X-axis is the left-right direction of the vehicle 20, and the Z-axis is the height direction of the vehicle 20.

[0138] In addition, Figures 7A to 7C is used to explain the case of calculating the center of gravity position of the vehicle 20 when a passenger is sitting in the front row seat and the left side of the rear row seat. The vehicle center w1 represents the center of gravity position of the vehicle 20 when no one is in the vehicle. The passenger center of gravity w2 represents a passenger sitting on the left side of the front row seat (for example, Figure 2The center-of-gravity position of the shown user 1). The passenger center of gravity w3 represents a passenger sitting on the left side of the rear seat (e.g., Figure 2 The center-of-gravity position of the shown user 2).

[0139] As Figures 7A to 7C shown, the passenger center of gravity w2 and the passenger center of gravity w3 are respectively represented as (lx2, ly2) and (lx3, ly3) on the XY plane with the vehicle center w1 as the origin.

[0140] In addition, when the weight of the vehicle 20 is set as m1, the weight of the passenger sitting on the left side of the front seat is set as m2, and the weight of the passenger sitting on the left side of the rear seat is set as m3, the center-of-gravity position (lx, ly) of the vehicle 20 when two passengers are on board can be calculated by the following formula 1 and formula 2.

[0141] lx = (lx2 × m2 + lx3 × m3) / (m1 + m2 + m3) (Formula 1)

[0142] ly = (ly2 × m2 + ly3 × m3) / (m1 + m2 + m3) (Formula 2)

[0143] The modeling unit 12 calculates the above center-of-gravity position (lx, ly) according to each past reservation information 14a of the vehicle 20 by using Formula 1 and Formula 2. Thus, for example, for the vehicle 20 with the same vehicle ID, multiple center-of-gravity positions (lx, ly) are calculated.

[0144] Referring again to Figure 6 , the modeling unit 12 performs regression analysis processing by using the calculated center-of-gravity position (lx, ly) and the obtained estimated value of the steering angle center, thereby generating the model 14d (S14). When the modeling unit 12 sets lx in the calculated center-of-gravity position as the X axis, ly in the center-of-gravity position as the y axis, and the steering angle center value δ at that time as the z axis, it plots multiple points (point group) on the xyz coordinates, and performs plane fitting by using the least squares method, thereby calculating the coefficients a, b, and c of the following formula 3. In other words, the center-of-gravity position (lx, ly) and the steering angle center value δ are used as the input information for generating the model 14d.

[0145] a × lx + b × ly + c = δ (Formula 3)

[0146] Regarding the rudder angle center value δ, for example, the value of the estimated rudder angle center of the driving information 14c is used. Thus, the modeling unit 12 can calculate the coefficients a, b, and c of the vehicle 20 suitable for the vehicle ID for each vehicle ID, for example. Calculating the coefficients a, b, and c through regression analysis (for example, linear regression analysis) in this way is an example of generating the model 14d. In addition, Equation 3 is an example of a predefined basic model. Step S14 is a step of generating the model 14d, and the model 14d is generated using the predefined basic model.

[0147] In this way, the model 14d is generated, for example, based on the center of gravity position (lx, ly) of the vehicle 20, the estimated value of the rudder angle center, and the basic model. The center of gravity position of the vehicle 20 is calculated based on the past reservation information 14a and the vehicle information 14b.

[0148] Next, the modeling unit 12 determines whether the regression analysis shrinks (S15). The modeling unit 12 determines whether the error of the least squares method is below a certain value, for example, to determine whether the regression analysis shrinks. For example, the modeling unit 12 determines whether the total value of the squares of the distances between each point group and the plane defined by the coefficients a, b, and c is below a specified value. The specified value is a value set in advance.

[0149] When the regression analysis shrinks (\"Yes\" in S15), the modeling unit 12 stores the model 14d generated in step S14 in the information management unit 14 (S16). In other words, the modeling unit 12 stores the coefficients a, b, and c shown in Equation 3 in the information management unit 14.

[0150] In addition, when the number of samples is below the first threshold (\"No\" in S12) and when the regression analysis does not shrink (\"No\" in S15), the modeling unit 12 returns to step S11 to continue the process.

[0151] In addition, the method for calculating the coefficients a, b, and c by the modeling unit 12 is not limited to using the least squares method, and any existing method can be used. In addition, the basic model is not limited to the model represented by Equation 3, and any model used in regression analysis can be used.

[0152] Next, regarding the prediction process, refer to Figure 8 for the description. Figure 8 is a flowchart showing the details of the prediction process (S20) of the present embodiment. Figure 8 The process shown is executed by the first prediction unit 13, for example. Figure 8 The process shown is executed after the current reservation information 14a is obtained via the communication unit 11.

[0153] As Figure 8As shown, in the prediction process (S20), the first prediction unit 13 reads the model 14d from the information management unit 14 (S21) and determines whether the model 14d exists (S22). The first prediction unit 13, for example, determines whether there is a model 14d suitable for the vehicle 20. Specifically, the first prediction unit 13, for example, determines whether there is a model 14d of the vehicle 20 with a specified vehicle ID. The specified vehicle ID is, for example, the vehicle ID included in the current reservation information 14a obtained via the communication unit 11. In other words, the specified vehicle ID is the vehicle ID of the vehicle 20 for which the predicted rudder angle center value is to be predicted.

[0154] When the model 14d exists (Yes in S22), the first prediction unit 13 reads out the model 14d from the information management unit 14. Specifically, the coefficients a, b, and c in Equation 3 as the basic model are read out from the information management unit 14.

[0155] Next, the first prediction unit 13 reads the reservation information 14a (S23). The first prediction unit 13, for example, reads out the current reservation information 14a from the information management unit 14. Step S23 is a step for obtaining the layout of the passengers and the weights of the passengers in the vehicle 20. Step S23 is an example of obtaining the first passenger information.

[0156] Next, the first prediction unit 13 reads the vehicle information 14b (S24). The first prediction unit 13, for example, reads out the vehicle information 14b of the vehicle ID included in the current reservation information 14a from the information management unit 14. Step S24 is a step for obtaining the weight of the vehicle 20. Step S24 is an example of obtaining the weight of the vehicle 20.

[0157] Next, the first prediction unit 13 uses the read model 14d to predict the predicted rudder angle center value of the vehicle 20 (S25). The first prediction unit 13 calculates the center of gravity position (lx, ly) according to the current reservation information 14a and Equations 1 and 2, and substitutes the calculated center of gravity position (lx, ly) into Equation 3. Thereby, the rudder angle center value δ is calculated. The rudder angle center value δ calculated in step S25 is an example of the predicted rudder angle center value. In addition, step S25 is an example of generating the predicted rudder angle center value.

[0158] Next, the first prediction unit 13 determines whether the predicted rudder angle center value predicted in step S25 is within a specified range (S26). Regarding the specified range, as long as the range is within a range where it is possible to determine whether the predicted rudder angle center value actually exists, the specified range is set in advance. When the predicted rudder angle center value is within the specified range ( "Yes" in S26), the first prediction unit 13 outputs the predicted predicted rudder angle center value to the vehicle 20 (S27). For example, the first prediction unit 13 outputs the predicted rudder angle center value of the steering center predicted in step S25 to the vehicle 20 determined by the vehicle ID included in the current reservation information 14a. Step S27 is an example of outputting information indicating the predicted rudder angle center value to the vehicle 20.

[0159] In addition, when the model 14d does not exist ( "No" in S22) and when the predicted rudder angle center value is outside the specified range ( "No" in S26), the first prediction unit 13 returns to step S21 to continue the process.

[0160] In addition, the first prediction unit 13 may execute the prediction process (S20) before the vehicle 20 travels under the current reservation information 14a. For example, the first prediction unit 13 may execute the prediction process (S20) before the "time" of the current reservation information 14a elapses.

[0161] Next, regarding the dynamic estimation process, refer to Figure 9 and Figure 10 for description. Figure 9 FIG. is a flowchart showing the details of the dynamic estimation process (S30) of the present embodiment. Figure 9 The process shown is executed, for example, by the dynamic estimation unit 22a of the vehicle 20. Figure 9 The process shown is executed after the vehicle 20 obtains the predicted rudder angle center value from the prediction device 10.

[0162] As Figure 9As shown, in the dynamic estimation process (S30), the dynamic estimation unit 22a reads the predicted steering angle center value of the steering center from a storage device (not shown) provided in the vehicle 20. This predicted steering angle center value is a value obtained from the prediction device 10 and stored in this storage device (S31), and determines whether there is a predicted steering angle center value (S32). When there is a predicted steering angle center value (Yes in S32), the dynamic estimation unit 22a reads the predicted steering angle center value and sets the read predicted steering angle center value as the initial value of the dynamic estimation process (S33). When there are multiple predicted steering angle center values, the dynamic estimation unit 22a sets the latest predicted steering angle center value as the initial value of the dynamic estimation process. In addition, when there is no predicted steering angle center value (No in S32), the dynamic estimation unit 22a returns to step S31 to continue the process.

[0163] Next, the dynamic estimation unit 22a determines whether the vehicle 20 has started to move based on the detection results of various sensors (S34). The dynamic estimation unit 22a determines whether the vehicle 20 has started to move based on the detection results of the vehicle speed sensor 23c, for example. And when the vehicle 20 has started to move (Yes in S34), the dynamic estimation unit 22a starts the operation of the Kalman filter in order to estimate the steering angle center value during the movement of the vehicle 20 in real time (S35). In other words, the dynamic estimation unit 22a estimates the steering angle center value during the movement of the vehicle itself through the estimation process using the Kalman filter. The operation of the Kalman filter will be described later. In addition, when the vehicle 20 has not started to move (No in S34), the dynamic estimation unit 22a returns to step S34 to continue the process.

[0164] Next, the dynamic estimation unit 22a determines whether the estimation variance (variance value) is less than the second threshold (S36). The dynamic estimation unit 22a calculates the variance (variance value) of the steering angle center value estimated using the Kalman filter and determines whether the calculated variance value is less than the second threshold. This variance value corresponds to the variance value P calculated by Equation 13 described later. And when the variance value P is smaller than the second threshold (Yes in S36), the dynamic estimation unit 22a reflects the estimated steering angle center value in the steering control (S37). Specifically, the dynamic estimation unit 22a outputs the estimated steering angle center value to the steering control unit 22b. The steering control unit 22b controls the steering angle of the vehicle itself using the steering angle center value estimated by the dynamic estimation unit 22a.

[0165] Here, regarding the operation of the Kalman filter, etc., reference Figure 10 will be described. Figure 10This is a diagram for explaining the operation of the extended Kalman filter according to the present embodiment. In addition, the extended Kalman filter according to the present embodiment is a filter for estimating the deviation of the steering angle (steering angle deviation δb).

[0166] As Figure 10 shown, the dynamic estimation unit 22a has, for example, a second prediction unit 22a1, an observation update unit 22a2, and an arithmetic unit 22a3.

[0167] The second prediction unit 22a1 predicts the state of the vehicle 20 based on the steering angle δ from the steering angle sensor 23b and the vehicle speed vx in the traveling direction from the vehicle speed sensor 23c. The second prediction unit 22a1, for example, sets the predicted value as Mathematical Formula 1,

[0168] [Mathematical Formula 1]

[0169]

[0170] When the state of the vehicle 20 is set as x, the following Formula 4 is used for prediction.

[0171] [Mathematical Formula 2]

[0172]

[0173] In addition, in the present embodiment, there is no control input, so the term of the control input is omitted in Formula 4. As shown in Formula 4, without the term of the control input in the present embodiment, the next information can be predicted by simple calculation.

[0174] Here, for the predicted value

[0175] [Mathematical Formula 3]

[0176]

[0177] The model for making the prediction is expressed as the following Formula 5, and the state x is expressed as the following Formula 6.

[0178] [Mathematical Formula 4]

[0179]

[0180] Mathematical Formula 4 represents the acceleration of the vehicle 20, δ represents the steering angle detected by the steering angle sensor 23b, and δb represents the steering angle deviation. For example, the predicted steering angle center value predicted by the first prediction unit 13 of the prediction device 10 is used as the initial value of the steering angle deviation δb.

[0181] [Mathematical Formula 5]

[0182]

[0183] [Mathematical Formula 6]

[0184]

[0185] In addition, when the yaw angular velocity is set to γ, the observed value z is calculated by the following Equation 7.

[0186] [Equation 7]

[0187]

[0188] Furthermore, the second prediction unit 22a1 further calculates the variance value (estimated variance value) shown in Equation 8, where Equation 8 is the variance value.

[0189] [Equation 8]

[0190]

[0191] Specifically, the variance value is Equation 9,

[0192] [Equation 9]

[0193]

[0194] When the Jacobian determinant of the prediction model is set to F, the variance value is set to P, and the process noise is set to Q, it is calculated by the following Equation 8.

[0195] [Equation 10]

[0196]

[0197] Here, the Jacobian determinant F of the prediction model is defined, for example, by the following Equation 9.

[0198] [Equation 11]

[0199]

[0200] Furthermore, the variance value P is updated sequentially by the observation update unit 22a2, for example. The process noise Q is obtained in advance.

[0201] The second prediction unit 22a1 predicts the state

[0202] [Equation 12]

[0203]

[0204] and the variance value

[0205] [Equation 13]

[0206]

[0207] Output to the observation update unit 22a2.

[0208] The observation update unit 22a2 uses the obtained state

[0209] [Mathematical formula 14]

[0210]

[0211] and the variance value

[0212] [Mathematical formula 15]

[0213]

[0214] to calculate the Kalman gain K. The observation update unit 22a2 calculates the Kalman gain K using, for example, Equation 10 below.

[0215] [Mathematical formula 16]

[0216]

[0217] Here, R is the variance value of the observation error. In addition, H is the Jacobian determinant of the observation model (Mathematical formula 17 described later),

[0218] [Mathematical formula 17]

[0219]

[0220] and is defined, for example, by Equation 11 below.

[0221] [Mathematical formula 18]

[0222]

[0223] Here, L is the length L between the front axle and the rear axle of the front wheels (refer to Figure 7A ).

[0224] Next, the observation update unit 22a2 updates the state x and the variance value P using the observation value z. The estimated value x is calculated by Equation 12 below, and the variance value P is calculated by Equation 13 below.

[0225] [Mathematical formula 19]

[0226]

[0227] [Mathematical formula 20]

[0228]

[0229] Here, h(x) is the observation model and is calculated by Equation 14 below.

[0230] [Mathematical formula 21]

[0231]

[0232] In addition, the mathematical formula in the third line of Equation 14 is a mathematical formula based on the relationship between the rudder angle δ and the yaw angular velocity γ, which connects the observation model h(x) with the yaw angular velocity γ.

[0233] In addition, the calculation of the mathematical formula 22 shown in Equation 12

[0234] [Mathematical formula 22]

[0235]

[0236] is equivalent to estimating the difference between the predicted value and the actual yaw angular velocity γ.

[0237] Moreover, the arithmetic unit 22a3 calculates the rudder angle δ and the rudder angle deviation δb to estimate the center value of the rudder angle, and outputs the estimated center value of the rudder angle to the steering control unit 22b. The calculation can be, for example, an addition operation, or a weighted addition operation, etc. In addition, the calculation can also be a subtraction operation, a multiplication operation, a division operation, etc.

[0238] In addition, the second prediction unit 22a1 predicts the center value of the rudder angle at the next moment based on the estimated value x and the variance value P estimated by the observation update unit 22a2. Such processing is repeatedly executed.

[0239] (Modification Example 1 of the Embodiment)

[0240] The following describes the information processing system 1a related to this modification example with reference to Figure 11 and Figure 12 for explanation. Figure 11 is a block diagram showing the functional structure of the information processing system 1a of this modification example. The information processing system 1a of this modification example is mainly different from the information processing system 1 of the embodiment in that the prediction device 10a has a detection unit 115, and the sensor unit 123 of the vehicle 20a has a human detection sensor 123d. Next, the differences between the information processing system 1a of this modification example and the information processing system 1 of the embodiment will be mainly described. In addition, in this modification example, the same or similar structures as those of the information processing system 1 of the embodiment are given the same reference numerals as those of the information processing system 1, and the description is omitted or simplified.

[0241] As Figure 11As shown, the information processing system 1a includes a prediction device 10a and a vehicle 20a. In the prediction device 10a, in addition to the prediction device 10 of the embodiment, a detection unit 115 is further included. In addition, instead of the sensor unit 23 of the embodiment, the sensor unit 123 included in the vehicle 20a has a person detection sensor 123d. In the embodiment, the predicted rudder angle center value is predicted using the current reservation information 14a. However, when actually moving based on this reservation information 14a, sometimes the passengers do not sit at the sitting positions according to this reservation information 14a. Thus, in this modified example, when actually moving based on this reservation information 14a, the information processing system 1a predicts the predicted rudder angle center value according to the actual sitting positions of the passengers.

[0242] The person detection sensor 123d detects the passengers sitting in the vehicle 20. The person detection sensor 123d is not particularly limited as long as it can detect the positions of the passengers sitting in the vehicle 20. For example, it can be implemented by a weight sensor, a camera sensor, etc. In this modified example, the person detection sensor 123d is a pressure sensor provided at each seat. The person detection sensor 123d outputs the detected pressure value to the prediction device 10a via the communication unit 21. The pressure value is an example of the output data. In addition, the person detection sensor 123d is an example of a transported object detection sensor that detects the transported objects carried in the vehicle 20.

[0243] The detection unit 115 generates the layout of the passengers using the output data obtained from the vehicle 20a. Specifically, the detection unit 115 determines whether there are passengers sitting at each seat in the vehicle 20a according to the pressure value obtained from the vehicle 20a. In addition, the detection unit 115 generates the weights of the passengers using the output data obtained from the vehicle 20a. Specifically, the detection unit 115 calculates the weights of the passengers sitting at each seat in the vehicle 20a according to the pressure value obtained from the vehicle 20a. Additionally, when the person detection sensor 123d is a camera, the detection unit 115 can determine whether a passenger is sitting by image analysis.

[0244] When there is a difference between the current reservation information 14a and the judgment result of the detection unit 115, in other words, when the sitting position in this reservation information 14a is different from the actual sitting position, the first prediction unit 13 predicts the predicted rudder angle center value again. It can be said that the first prediction unit 13 changes the predicted rudder angle center value.

[0245] In addition, when the first prediction unit 13 obtains a pressure value from the human detection sensor 123d, it may not perform the process of estimating the predicted rudder angle center value using the current reservation information 14a. The first prediction unit 13 may use the determination result of the detection unit 115 as the first passenger information to estimate the predicted rudder angle center value. In other words, the first passenger information may also be generated using the output data of the human detection sensor 123d provided in the vehicle 20.

[0246] As described above, the information processing system 1a according to this modification includes: a vehicle 20a having a human detection sensor 123d, and a detection unit 115 that detects a passenger based on the detection result of the human detection sensor 123d. Moreover, the first prediction unit 13 predicts the predicted rudder angle center value using the determination result of the detection unit 115.

[0247] Thus, the information processing system 1a can predict the predicted rudder angle center value corresponding to the actual seating position in the vehicle 20. The information processing system 1a can also predict the predicted rudder angle center value, for example, when the current reservation information 14a cannot be obtained or the current reservation information 14a does not include seat information.

[0248] Next, regarding the operation of the above-described information processing system 1a, refer to Figure 12 for description. Figure 12 FIG. is a flowchart showing the operation of the information processing system 1a of this modification. Figure 12 The process shown is performed, for example, between Figure 5 the steps S20 and S30 shown, but may also be executed, for example, instead of step S23. In addition, in Figure 12 , the prediction device 10a is set to obtain a pressure value from the human detection sensor 123d and store the pressure value in the information management unit 14.

[0249] Figure 12 As shown, the detection unit 115 reads the pressure value from the information management unit 14 (S41), and determines whether the pressure value is greater than a third threshold value (S42). The third threshold value may be any value that can determine whether a passenger is seated on the seat. When the pressure value is greater than the third threshold value (Yes in S42), the detection unit 115 updates the seating information in the vehicle 20 assuming that a person (passenger) is seated on the seat (S43). In addition, when the pressure value is less than or equal to the third threshold value (No in S42), the detection unit 115 updates the seating information assuming that a person (passenger) is not seated on the seat (S44).

[0250] Next, the detection unit 115 determines whether the determination in step S42 has been made for all seats (S45). When the detection unit 115 has made the determination in step S42 for all seats ( "Yes" in S45), it proceeds to step S46. When the determination in step S42 has not been made for all seats ( "No" in S45), it returns to step S41 to continue processing the seats.

[0251] When there is current reservation information 14a, the detection unit 115 determines whether the seating position in the current reservation information 14a is different from the actual seating position based on the seating information (S46). Regarding differences in the seating position, examples include different seating positions and different numbers of passengers. In addition, when the person detection sensor 123d is a camera and the detection unit 115 determines the person on board based on image analysis, the detection unit 115 can also determine whether the person determined is the person included in the current reservation information 14a. The detection unit 115 outputs the determination result to the first prediction unit 13.

[0252] When the first prediction unit 13 obtains a determination result indicating that the seating position in the current reservation information 14a is different from the actual seating position based on the seating information, it uses the seating information to predict the predicted rudder angle center value again (S47). In other words, the first prediction unit 13 predicts the predicted rudder angle center value using the actual seating position of the passengers. The process of predicting the predicted rudder angle center value is the same as the prediction process (S20), so the description is omitted. The first prediction unit 13 outputs the re - predicted predicted rudder angle center value to the vehicle 20a. In addition, when there is no current reservation information 14a, the process of step S46 is not performed, and the process of step S47 is executed.

[0253] Thus, the vehicle 20a can obtain a predicted rudder angle center value corresponding to the actual seating condition (e.g., seating position), so it can control the rudder angle corresponding to the actual seating condition.

[0254] In addition, in this modified example, an example is described in which the predicted rudder angle center value is predicted using the current reservation information 14a and then the predicted rudder angle center value is changed corresponding to the output data of the person detection sensor 123d. However, it is not limited to this. When the prediction device 10a obtains the output data of the person detection sensor 123d, it may not perform the prediction of the predicted rudder angle center value using the current reservation information 14a. In other words, the prediction device 10a can use the output data of the person detection sensor 123d to predict the predicted rudder angle center value.

[0255] (Modified Example 2 of the Embodiment)

[0256] Next, regarding the information processing system 1b of this modified example, refer toFigures 13 to 16 will be described. Figure 13 Figure 13 is a block diagram showing the functional structure of the information processing system 1b of this modification example. The information processing system 1b of this modification example is mainly different from the information processing system 1 of the embodiment in that the modeling unit 212 of the prediction device 10b generates a machine learning model. Next, the differences between the information processing system 1b of this modification example and the information processing system 1 of the embodiment will be described centering on the differences. In addition, in this modification example, for the structures that are the same as or similar to those of the information processing system 1 of the embodiment, the same reference numerals as those of the information processing system 1 are given, and the description is omitted or simplified.

[0257] As Figure 13 shown, the information processing system 1b includes a prediction device 10b and a vehicle 20. The prediction device 10b has a communication unit 11, a modeling unit 212, a first prediction unit 13, and an information management unit 214.

[0258] The modeling unit 212 generates a machine learning model, that is, a model 214d. The modeling unit 212 generates a model 214d that takes the layout and weight of the passengers and the weight of the vehicle 20 as inputs and outputs the predicted rudder angle center value at that time. The model 214d is, for example, a neural network type learning model, but is not limited thereto.

[0259] When the first prediction unit 13 obtains the current reservation information 14a, it reads the model 214d and outputs the predicted rudder angle center value obtained by inputting the layout and weight of the passengers in the reservation information 14a and the weight of the vehicle 20 to the model 214d read to the vehicle 20.

[0260] The information management unit 214 further stores various information for generating the model 214d on the basis of the information management unit 14 of the embodiment. The information management unit 214 further stores, for example, user information 214e and environment information 214f.

[0261] The user information 214e includes the attribute information of the passengers. The user information 214e includes, for example, Figure 14 shown "name", "ID", "gender", "age", and "place of birth". Figure 14 is a diagram showing an example of the user information 214e of this modification example. In addition, "gender", "age", and "place of birth" are examples of the attribute information of the passengers. The attribute information of the passengers includes, for example, information that may be related to the physical characteristics of the passengers.

[0262] "Name" represents the name of the passenger, and may also be a nickname, for example.

[0263] "ID" is identification information for identifying the passenger, for example, an identification number set for each passenger.

[0264] "Gender" indicates the gender of the passenger.

[0265] "Age" indicates the age of the passenger. "Age" can be a specific numerical value or an age range.

[0266] "Place of birth" indicates the place of birth of the passenger.

[0267] In the user information 214e, for example, for each of multiple vehicles 20 (each vehicle from vehicle 1 to vehicle 4 in Figure 14 ), the above items are set respectively. Taking "vehicle 1" as an example, information on the above items is stored for users A, F, and N respectively. A, F, and N can be people who have ridden in "vehicle 1" or reserved "vehicle 1" in the past, or people who will make a reservation in the future. Such information is pre-registered information.

[0268] In addition, the user information 214e may further include other information for estimating the weight of the passenger. The user information 214e can include, for example, height, body type, or the weight itself.

[0269] The environmental information 214f includes information related to the external environment of the vehicle 20. The environmental information 214f, for example, includes Figure 15 the "temperature", "humidity", "road surface condition", "wind direction", and "wind speed" shown. Figure 15 It is a diagram showing an example of the environmental information 214f of this modified example.

[0270] "Temperature" and "humidity" indicate the temperature and humidity around the vehicle 20 is traveling.

[0271] "Road surface condition" indicates the condition of the road surface on which the vehicle 20 is traveling.

[0272] "Wind direction" and "wind speed" indicate the wind direction and wind speed around the vehicle 20 is traveling.

[0273] When the prediction device 10b obtains the planned driving route based on the "departure place" and "destination" in the current reservation information 14a, it can obtain the environmental information around the driving route from a server device that manages environmental information, etc. In addition, the environmental information 214f is an example of external environmental information.

[0274] As described above, the information processing system 1b according to this modified example has a modeling unit 212 that generates the model 214d through machine learning.

[0275] Thus, the information processing system 1b can predict the predicted rudder angle center value with higher accuracy by using a machine learning model (model 214d) instead of the least squares method. When generating the machine learning model, the information processing system 1b can generate the model 214d that can predict the predicted rudder angle center value with further higher accuracy by using the user information 214e and the environment information 214f. In addition, in order to reduce the dependence of the machine learning model on the vehicle 20, the information processing system 1b can estimate the center of gravity position of the vehicle 20 with the machine learning model and calculate the relationship between the center of gravity position and the predicted rudder angle center value by using the least squares method.

[0276] Next, regarding the operation of the above-described information processing system 1b, reference will be made to Figure 16 explain. Figure 16 is a flowchart showing the operation of the information processing system 1b of this modification example. Figure 16 The process shown is a modeling process (S10a). For example, instead of Figure 4 shown in step S10, step S10a is executed.

[0277] As Figure 16 shown, the modeling unit 212 reads various data for generating the model 214d from the information management unit 214 (S51). The modeling unit 212 reads at least one of, for example, the past reservation information 14a, the vehicle information 14b, the driving information 14c, the user information 214e, and the environment information 214f as training data for generating the model 214d. For example, the training data may include the user information 214e of the passenger. In other words, the training data may include the attribute information of the passenger. In addition, for example, the training data may also include the environment information 214f outside the vehicle 20. The modeling unit 212 can read all of the past reservation information 14a, the vehicle information 14b, the driving information 14c, the user information 214e, and the environment information 214f as training data, for example.

[0278] Next, the modeling unit 212 determines whether the number of samples of the read various data is greater than the fourth threshold (S52). In step S52, the modeling unit 212 determines whether the number of data samples is sufficient to generate the model 214d. The fourth threshold is not particularly limited as long as it is the number of samples with which the modeling unit 212 can generate the model 214d. By making the determination in step S52, the modeling unit 212 can generate the model 214d by using various data in various situations.

[0279] The modeling unit 212 calculates an estimated value of the rudder angle center when the number of samples is greater than the fourth threshold ("Yes" in S52) (S53). The processing of step S53 is the same as that of step S13, so the description is omitted. In addition, when the number of samples is less than or equal to the fourth threshold ("No" in S52), the modeling unit 212 returns to step S51 to continue the processing.

[0280] In the construction of the model 214d by the modeling unit 212, various methods can be used. In the present embodiment, the modeling unit 212 uses an ensemble learning method of combining multiple weak learners to construct the model 214d, that is, referring to the learning result of one weak learner to learn the next weak learner, and using the gradient boosting method of the gradient of the loss function to construct the model 214d (S54 and S55). The loss function is a function that outputs the difference between the predicted value and the correct value (in other words, the correct answer value).

[0281] In addition, the construction of the model 214d is not limited to the above method. For example, it can be an ensemble learning such as Bagging, or other methods.

[0282] In this way, the model 214d is, for example, a model generated by performing machine learning using past reservation information 14a and vehicle information 14b as training data and the rudder angle center value as correct answer data.

[0283] Next, the modeling unit 212 determines whether the generated model 214d shrinks (S56). When the model 214d shrinks ("Yes" in S56), the modeling unit 212 stores the generated model 214d in the information management unit 214 (S57). In addition, when the model 214d does not shrink ("No" in S56), the modeling unit 212 returns to step S51 to continue the processing.

[0284] (Modification Example 3 of the Embodiment)

[0285] Next, regarding the information processing system 1c of this modification example, refer to Figures 17 to 19 for the description. Figure 17 is a block diagram showing the functional structure of the information processing system 1c of this modification example. The information processing system 1c of this modification example is mainly different from the information processing system 1 of the embodiment in that the vehicle 20c is provided with a determination unit 325. Next, the differences between the information processing system 1c of this modification example and the information processing system 1 of the embodiment will be mainly described. In addition, in this modification example, the same or similar structures as those of the information processing system 1 of the embodiment are given the same reference numerals as those of the information processing system 1, and the description is omitted or simplified.

[0286] As Figure 17As shown, the information processing system 1c includes a prediction device 10 and a vehicle 20c. In addition to the vehicle 20 of the embodiment, the vehicle 20c has a determination unit 325 that determines whether the vehicle 20c is moving straight ahead. This is because the estimation accuracy of the dynamic estimation process is higher when the vehicle 20c is moving straight ahead than when it is turning.

[0287] Based on the vehicle 20 of the embodiment, the vehicle 20c further has a determination unit 325 that determines whether the vehicle 20 is moving straight ahead according to the output data of the sensor unit 23. In addition, instead of the dynamic estimation unit 22a, the vehicle 20c has a dynamic estimation unit 322a.

[0288] For example, the determination unit 325 determines whether the vehicle 20c is moving straight ahead by using the output data of the gyro sensor 23a, i.e., the yaw angular velocity.

[0289] Based on the determination result of the determination unit 325, the dynamic estimation unit 322a performs dynamic estimation processing. Specifically, the dynamic estimation unit 322a performs dynamic estimation processing when the vehicle itself is moving straight ahead. For example, the dynamic estimation unit 322a performs dynamic estimation processing only when the vehicle itself is moving straight ahead. In other words, the dynamic estimation unit 322a does not perform dynamic estimation processing when the vehicle itself is turning.

[0290] As described above, the information processing system 1c according to this modification includes a determination unit 325 that determines whether the vehicle itself is moving straight ahead. Moreover, when the determination unit 325 determines that the vehicle itself is moving straight ahead, the dynamic estimation unit 322a performs dynamic estimation processing.

[0291] Thus, the information processing system 1c can perform dynamic estimation of the steering angle center only when moving straight ahead, so the estimation accuracy can be improved.

[0292] Next, regarding the operation of the above information processing system 1c, reference is made to Figure 18 and Figure 19 for description. Figure 18 This is a flowchart showing the determination process of this modification. In Figure 18 it is set that the output data of the gyro sensor 23a, i.e., the gyro sensor value, is sequentially stored in the storage device (not shown) of the vehicle 20c. The gyro sensor value is, for example, the yaw angular velocity.

[0293] As Figure 18As shown, the determination unit 325 reads the gyro sensor value (S61). The determination unit 325, for example, reads the latest yaw angular velocity from the storage device. Moreover, the determination unit 325 determines whether the absolute value of the yaw angular velocity is greater than the fifth threshold value (S62). The fifth threshold value may be any value that can determine whether the vehicle itself is moving straight forward, and is not particularly limited.

[0294] When the absolute value of the yaw angular velocity is greater than the fifth threshold value (Yes in S62), the determination unit 325 determines that the vehicle itself is not moving straight forward. In other words, the vehicle itself is turning (S63). When the absolute value of the yaw angular velocity is less than or equal to the fifth threshold value (No in S62), the determination unit 325 determines that the vehicle itself is moving straight forward (S64). Moreover, the determination unit 325 outputs the determination result to the dynamic estimation unit 322a (S65).

[0295] The determination unit 325 can perform the Figure 18 actions shown at regular time intervals, or can perform them sequentially.

[0296] Next, regarding the operation of the dynamic estimation unit 322a, reference is made to Figure 19 for explanation. Figure 19 is a flowchart showing the details of the dynamic estimation process (S30a) of this modified example.

[0297] As Figure 19 shown, the dynamic estimation process (S30a) of this modified example further includes steps S71 and S72 on top of the dynamic estimation process (S30) of the embodiment.

[0298] When the dynamic estimation unit 322a obtains the determination result from the determination unit 325 (S71), it determines whether the vehicle itself is moving straight forward based on the determination result (S72). When the determination result includes information indicating that the vehicle itself is moving straight forward, in other words, when the vehicle itself is moving straight forward (Yes in S72), the dynamic estimation unit 322a performs the processing after step S31. In addition, when the determination result includes information indicating that the vehicle itself is not moving straight forward, in other words, when the vehicle itself is not moving straight forward (No in S72), the dynamic estimation unit 322a returns to step S71 and continues the processing. In other words, when it is No in step S72, the estimation process in the dynamic estimation process is not performed.

[0299] (Modified Example 4 of the Embodiment)

[0300] Next, regarding the information processing system 1d related to this modified example, reference is made to Figure 20 and Figure 21 for explanation. Figure 20It is a block diagram showing the functional structure of the information processing system 1d of this modified example. The information processing system 1d of this modified example is mainly different from the information processing system 1 of the embodiment in that the sensor unit 423 of the vehicle 20d has a camera 423e and a passenger estimation unit 416 that estimates a passenger based on the image captured by the camera 423e. Next, the differences between the information processing system 1d of this modified example and the information processing system 1 of the embodiment will be mainly described. In addition, in this modified example, for the structures that are the same as or similar to those of the information processing system 1 of the embodiment, the same reference numerals as those of the information processing system 1 are given, and the description is omitted or simplified.

[0301] As Figure 20 shown, the information processing system 1d includes a prediction device 10d and a vehicle 20d. In the information processing system 1d of this modified example, the passenger estimation unit 416 of the prediction device 10d estimates the weight and the center of gravity position of the passenger based on the image, and the modeling unit 12 further estimates the center of gravity position of the vehicle 20d using the estimation result 414g.

[0302] The vehicle 20d has a sensor unit 423 including a camera 423e. The camera 423e captures the interior condition of the vehicle 20d. The camera 423e captures an image from which the weight and the center of gravity position of the passenger sitting in the vehicle 20d can be obtained. A plurality of cameras 423e may be provided. The vehicle 20d outputs the image captured by the camera 423e to the prediction device 10d.

[0303] The prediction device 10d has a passenger estimation unit 416 that estimates the weight and the center of gravity position of the passenger based on the image captured by the camera 423e, and the prediction device 10d also has an information management unit 414 that further stores the estimation result 414g of the passenger estimation unit 416.

[0304] The passenger estimation unit 416 estimates the weight of the passenger and the center of gravity position of the passenger by performing image analysis on the image captured by the camera 423e. The passenger estimation unit 416 estimates the weight and the center of gravity position of each passenger, for example. When estimating the weight and the center of gravity position, any existing technology can be used. When estimating the center of gravity position, for example, the posture of the passenger is estimated based on the image, and the estimated posture is used to estimate the center of gravity position.

[0305] The information management unit 414 accumulates a plurality of estimation results 414g. The estimation result 414g may also be stored in correspondence with the driving information 14c when the estimation result 414g is obtained.

[0306] The modeling unit 12 estimates the center of gravity position of the vehicle 20d using the estimation result 414g of the passenger estimation unit 416 in the modeling process.

[0307] As described above, the information processing system 1d of this modification example includes: a camera 423e that captures the interior of the vehicle itself, and a passenger estimation unit 416 that estimates the weight and the center of gravity position of a passenger sitting in the vehicle 20d by performing image analysis on the image captured by the camera 423e.

[0308] Thus, the information processing system 1d can correctly obtain the weight and the center of gravity position of the passenger, so that the estimation accuracy of the center of gravity position of the vehicle 20d can be improved. In addition, the modeling unit 12 can estimate the center of gravity position of the vehicle 20d even when the reservation information 14a has not been obtained in the past. In this case, the estimation result 414g is an example of the detection result. In addition, the passenger information is generated using the estimation result 414g. For example, the first passenger information is generated using the current estimation result 414g of the passenger based on the output data (image data) of the camera 423e in the vehicle 20, and the second passenger information is generated using the past estimation result 414g of the passenger based on the output data obtained in the past.

[0309] Next, regarding the operation of the above information processing system 1d, reference is made to Figure 21 for description. Figure 21 It is a flowchart showing the determination process of this modification example.

[0310] As Figure 21 shown, the passenger estimation unit 416 obtains the image data captured by the camera 423e (S81). The passenger estimation unit 416 can read the image data from the information management unit 414. Next, the passenger estimation unit 416 determines whether a person is detected in the obtained image data (S82). The passenger estimation unit 416 determines whether a person is photographed in the image by performing image analysis on the image data. In other words, the passenger estimation unit 416 determines whether there is a passenger in the vehicle 20d by performing image analysis on the image data.

[0311] When the passenger estimation unit 416 detects a person (Yes in S82), it estimates the weight and the center of gravity position of the person (S83). In other words, the passenger estimation unit 416 estimates the weight and the center of gravity position of the passenger. Then, the passenger estimation unit 416 stores the estimated estimation result 414g in the information management unit 414 (S84). In addition, when the passenger estimation unit 416 does not detect a person (No in S82), it returns to step S81 to continue the process.

[0312] In addition, when the passenger estimation unit 416 detects multiple people in step S82, the process of step S83 is executed for each of the multiple people.

[0313] In addition, the prediction device 10d can, on the basis of predicting the predicted rudder angle center value using the current reservation information 14a, further perform processing to change the predicted rudder angle center value according to the image data of the camera 423e. In addition, when the prediction device 10d obtains the image data from the camera 423e, it can refrain from predicting the predicted rudder angle center value using the current reservation information 14a. In other words, the prediction device 10d can use the image data from the camera 423e to predict the predicted rudder angle center value. The image data of the camera 423e can include information related to the layout of the passengers and the weights of the passengers.

[0314] (Modification Example 5 of the Embodiment)

[0315] Next, regarding the information processing system 1e related to this modification example, refer to Figure 22 and Figure 23 for description. Figure 22 is a block diagram showing the functional structure of the information processing system 1e of this modification example. The information processing system 1e of this modification example is mainly different from the information processing system 1d of Modification Example 4 of the embodiment in that the prediction device 10e has an authentication unit 517 instead of the passenger estimation unit 416. Next, the differences between the information processing system 1e of this modification example and the information processing system 1d of Modification Example 4 of the embodiment will be mainly described. In addition, in this modification example, for the structures that are the same as or similar to those of the information processing system 1d of Modification Example 4 of the embodiment, the same reference numerals as those of the information processing system 1d are given, and the description is omitted or simplified.

[0316] As Figure 22 shown, the information processing system 1e includes a prediction device 10e and a vehicle 20d. In the information processing system 1e related to this modification example, the authentication unit 517 of the prediction device 10e determines the individual passenger according to the authentication result of the image, and obtains the information related to the determined individual from the information management unit 514.

[0317] The authentication unit 517 determines the individual by performing face authentication on the image captured by the camera 423e, and outputs the authentication result to the modeling unit 12 and the first prediction unit 13. The authentication unit 517, for example, compares the authentication information 517h with the image data to determine the individual. In addition, the authentication unit 517 can store the authentication result in the information management unit 514. In addition, the method for determining the individual is not limited to face authentication. The authentication result is an example of the recognition result of the passenger. In addition, the recognition result includes the detection result of the passenger based on the image data of the camera 423e.

[0318] The modeling unit 12 obtains information corresponding to the passenger based on the authentication result, and estimates the center-of-gravity position of the vehicle 20d by using the obtained information. The modeling unit 12 generates the model 14d by using the estimated center-of-gravity position. The information corresponding to the passenger may be, for example, the user information 214e. In addition, the modeling unit 12 generates the model 14d by using, for example, the second passenger information, which is generated by using the user information 214e of the passenger determined according to the authentication result. Specifically, the modeling unit 12 generates the passenger information by using the attribute information in the user information 214e of the passenger determined according to the authentication result. The modeling unit 12 estimates the layout or weight of the passenger based on information such as "age" and "gender" included in the attribute information. For example, when the identified passenger is an elderly person, it is estimated that the passenger is sitting in the priority seat. The modeling unit 12 estimates the center-of-gravity position of the vehicle 20d by using the generated passenger information. The model 14d is generated by using the estimated center-of-gravity position.

[0319] The first prediction unit 13 obtains information corresponding to the passenger based on the authentication result, and estimates the center-of-gravity position of the vehicle 20d by using the obtained information. The first prediction unit 13 predicts the estimated rudder angle center value by using the estimated center-of-gravity position. The first prediction unit 13 predicts the estimated rudder angle center value by using, for example, the weight of the passenger obtained according to the authentication result. In addition, the first prediction unit 13 estimates the center-of-gravity position and predicts the estimated rudder angle center value by using, for example, the first passenger information generated by using the user information 214e of the passenger, where the passenger is determined according to the authentication result. Specifically, the first prediction unit 13 generates the passenger information by using the attribute information in the user information 214e of the passenger determined according to the authentication result. The generation process of the passenger information based on the attribute information is the same as the above process. The first prediction unit 13 estimates the center-of-gravity position of the vehicle 20d by using the generated passenger information. The estimated rudder angle center value is predicted by inputting the estimated center-of-gravity position into the model 14d.

[0320] In the information management unit 514, the authentication information 517h used for the authentication unit 517 to determine the passenger is stored. The authentication information 517h may be, for example, an image of the face of a registered user, or information representing the features of the user's face.

[0321] As described above, the information processing system 1e according to this modification example includes the camera 423e that captures the interior of the vehicle itself, and the authentication unit 517 that determines the passenger riding in the vehicle 20d by performing the authentication process on the image captured by the camera 423e.

[0322] Accordingly, the information processing system 1e can identify the person riding in the vehicle 20d. The modeling unit 12, for example, obtains the user information 214e of the identified person from the information management unit 514, thereby being able to correctly obtain information related to the person riding in the vehicle 20d. Thus, the modeling unit 12 can improve the estimation accuracy of the center of gravity position of the vehicle 20d, in other words, improve the estimation accuracy of the generated model 14d. In addition, the first prediction unit 13, for example, obtains the user information 214e of the identified person from the information management unit 514, thereby being able to correctly obtain information related to the person riding in the vehicle 20d. Thus, the first prediction unit 13 can improve the prediction accuracy of the predicted rudder angle center value of the vehicle 20d.

[0323] Next, regarding the operation of the above-described information processing system 1e, reference is made to Figure 23 for an explanation. Figure 23 It is a flowchart showing the determination process of this modification example.

[0324] As Figure 23 shown, the authentication unit 517 obtains the image data captured by the camera 423e (S91). The authentication unit 517 can read the image data from the information management unit 514. Next, the authentication unit 517 compares the analysis result of the obtained image data with the authentication information 517h stored in the information management unit 514 (S92) to determine whether it is a registered user (S93).

[0325] When the authentication unit 517 determines that it is a registered user (Yes in S93), it determines whether there is user information 214e corresponding to the user (S94). When the authentication unit 517 determines that there is user information 214e corresponding to the user (Yes in S94), it obtains the user information 214e (S95) and outputs the obtained user information 214e to at least one of the modeling unit 12 and the first prediction unit 13 (S95). In this modification example, the authentication unit 517 outputs the user information 214e to both the modeling unit 12 and the first prediction unit 13.

[0326] In addition, when the authentication unit 517 determines that there is no user information 214e corresponding to the user (No in S94), it outputs the registered user ID to at least one of the modeling unit 12 and the first prediction unit 13 (S97). Thus, the authentication unit 517 associates various data obtained later with the existing user ID and stores them in the information management unit 514.

[0327] In addition, when the user is not a registered user ("No" in S93), the authentication unit 517 registers the user as a new user (S98). For example, in step S98, the authentication unit 517 sets a new user ID for the user. Further, the authentication unit 517 outputs the set user ID to at least one of the modeling unit 12 and the first prediction unit 13 (S99). Thereby, the authentication unit 517 can associate various data obtained later with the newly registered user ID and store them in the information management unit 514.

[0328] (Other embodiments)

[0329] The present disclosure has been described above based on embodiments and modification examples (hereinafter also referred to as embodiments, etc.). However, the present disclosure is not limited to the above-described embodiments, etc. Within the scope not departing from the gist of the present disclosure, forms obtained by performing various modifications conceivable by those skilled in the art on the present embodiment and forms constructed by combining constituent elements in different embodiments are also included within the scope of one or more forms of the present disclosure.

[0330] For example, in the above-described embodiments, etc., an example in which the prediction device and the vehicle are separately configured has been described, but it is not limited thereto. For example, the prediction device may be mounted on the vehicle. In this case, as Figure 24 shown, the prediction device can function as a prediction unit of the vehicle. Figure 24 FIG. is a block diagram showing the functional configuration of an information processing system 1f according to another embodiment.

[0331] As Figure 24 shown, the information processing system 1f includes a vehicle 20f. The vehicle 20f includes a prediction unit 10f, a communication unit 21, a vehicle control unit 22, and a sensor unit 23. In addition, the control device 600 is composed of the first prediction unit 13, the communication unit 21, and the dynamic estimation unit 22a. In other words, the control device 600 is mounted on the vehicle 20.

[0332] In addition, in the above-described embodiments, etc., an example in which the reservation information includes the seating position of the passenger has been described, but it may further include the weight of each passenger, etc. Further, the reservation information may further include information related to the weight of the goods carried by the passenger.

[0333] In addition, in the above-described embodiments, etc., an example in which the transported object is a passenger has been described, but the transported object may also be goods mounted on the moving body.

[0334] In addition, in the above-described embodiments and the like, an example in which the reservation information is the reservation information of a passenger has been described. However, the reservation information may also be the reservation information of the goods carried on the moving body. The reservation information of the goods includes at least the loading location of the goods in the moving body and the weight of the goods. In addition, the reservation information of the goods may also include distribution plan information such as the collection time, delivery time, collection location, and delivery location of the goods, and information such as the type of the goods. For example, instead of the weight of the goods, the weight of the goods may be estimated or determined according to the type of the goods.

[0335] In addition, in the first modification of the above-described embodiment, an example in which the detection result is the detection result of a passenger has been described. However, the detection result may also be the detection result of the goods carried on the moving body. The detection result of the goods may include the detection results of the position, weight, etc. of the goods carried on the vehicle 20.

[0336] In addition, in the second modification of the above-described embodiment, an example in which the attribute information is the passenger attribute information has been described. However, the attribute information may be the attribute information of the goods carried on the moving body. The attribute information of the goods includes at least one of the "product name", "product number (such as product code)", "production date", and "place of origin" of the goods carried on the vehicle 20. For example, the attribute information of the goods may include information that may be related to the physical characteristics of the goods.

[0337] In addition, in the fifth modification of the above-described embodiment, an example in which the recognition result is the recognition result of a passenger has been described. However, the recognition result may also be the recognition result of the goods carried on the moving body. When an identification code (such as a QR code (registered trademark), etc.) for identifying the goods is attached to the surface of the goods, etc., the authentication unit determines the loaded goods based on the information obtained from the identification code of the image captured by the camera and the information stored in the information management unit for the authentication unit to determine the goods. The determined result is an example of the recognition result.

[0338] In addition, all or part of the information processing system of the above-described embodiments and the like may be implemented by a cloud server or may be implemented as an edge device carried on the moving body. For example, in the case where the moving body is an autonomous vehicle, the prediction device of the above-described embodiments and the like may also be implemented as a part of the autonomous driving device carried on the autonomous vehicle.

[0339] In addition, the order of the multiple processes described in the above-described embodiments and the like is an example. The order of the multiple processes may be changed, and the multiple processes may also be executed in parallel. In addition, a part of the multiple processes may not be executed.

[0340] In addition, each component described in the above embodiments and the like can be implemented as software, typically as an LSI of an integrated circuit. These can be made into individual chips, or can be made into a single chip in a way that includes part or all of them. Here, it is called an LSI, but depending on the degree of integration, it may sometimes be called an IC, a system LSI, a super LSI, or an ultra LSI. In addition, the method of integrating into an integrated circuit is not limited to LSI, and can be implemented using a dedicated circuit or a general-purpose processor. It is also possible to use an FPGA (Field Programmable Gate Array) that can be programmed after the LSI is manufactured, or a reconfigurable processor that can reconfigure the connection and setting of circuit units inside the reconfigurable LSI. Furthermore, as semiconductor technology advances or other derived technologies emerge, when there is an integrated circuit technology that can replace the LSI, of course, this technology can be used for the integration of components.

[0341] In addition, the division of functional blocks in the block diagram is an example. Multiple functional blocks can be implemented as one functional block, or one functional block can be divided into multiple ones, or part of the function can be transferred to other functional blocks. In addition, the functions of multiple functional blocks with similar functions can be processed by a single piece of hardware or software in parallel or time-sharing.

[0342] In addition, the prediction device included in the information processing system can be implemented as a single device, or can be implemented by multiple devices. For example, each processing unit of the prediction device can be implemented by two or more server devices. When the information processing system is implemented by multiple server devices, the components included in the information processing system can be allocated to the multiple server devices in any way. In addition, there is no particular limitation on the communication method between the multiple server devices. In addition, there is no particular limitation on the communication method between the components of the information processing system.

[0343] In addition, at least a part of the components of the vehicle described in the above embodiments and the like can be possessed by the prediction device. For example, the dynamic estimation unit can be possessed by the prediction device.

[0344] Furthermore, the technology of the present disclosure can be the above program, or can be a non-transitory computer-readable recording medium on which the above program is recorded. In addition, of course, the above program can be circulated via a transmission medium such as the Internet. For example, the above program and the digital signal constituted by the above program can be transmitted via a telecommunication line, a wireless or wired communication line, a network represented by the Internet, data broadcasting, etc. In addition, the above program and the digital signal constituted by the above program are recorded on a recording medium and transferred, or transferred via a network or the like, so that they can be executed by an independent other computer system.

[0345] In addition, in each of the embodiments, each component may be constituted by dedicated hardware or may be implemented by executing a software program suitable for each component. Each component may be implemented by a program execution unit such as a CPU or a processor that reads and executes a software program recorded in a recording medium such as a hard disk or a semiconductor memory.

[0346] Industrial Applicability

[0347] The present disclosure can be widely used in systems that utilize mobile bodies.

[0348] Symbol Explanation

[0349] 1, 1a, 1b, 1c, 1d, 1e, 1f Information Processing System

[0350] 10, 10a, 10b, 10d, 10e Prediction Device

[0351] 10f Prediction Unit

[0352] 11, 21 Communication Unit (First Acquisition Unit, Second Acquisition Unit, Output Unit)

[0353] 12, 212 Modeling Unit

[0354] 13 First Prediction Unit (Generation Unit)

[0355] 14, 214, 414, 514 Information Management Unit

[0356] 14a Reservation Information

[0357] 14b Vehicle Information

[0358] 14c Driving Information

[0359] 14d, 214d Model

[0360] 20, 20a, 20c, 20d, 20f Vehicle (Mobile Body)

[0361] 22 Vehicle Control Unit

[0362] 22a Dynamic Estimation Unit (Setting Unit)

[0363] 22a1 Second Prediction Unit

[0364] 22a2 Observation Update Unit

[0365] 22a3 Arithmetic Unit

[0366] 22b Steering Control Unit

[0367] 23, 123, 423 Sensor Unit

[0368] 23a Gyro sensor

[0369] 23b Rudder angle sensor

[0370] 23c Vehicle speed sensor

[0371] 115 Detection unit

[0372] 123d Person detection sensor

[0373] 214e User information

[0374] 214f Environmental information

[0375] 325 Judgment unit

[0376] 414 Information management unit

[0377] 414g Estimation result

[0378] 416 Passenger estimation unit

[0379] 423e Camera

[0380] 517 Authentication unit

[0381] 517h Authentication information

[0382] 600 Control device

[0383] a, b, c coefficients

[0384] lx, ly Center of gravity position

[0385] m1 Vehicle weight

[0386] m2, m3 Passenger weights

[0387] w1 Vehicle center

[0388] w2, w3 Passenger centers of gravity

[0389] δ Rudder angle center value

Claims

1. An information processing method, executed by a computer, in which the first conveyance information is obtained, where the first conveyance information is conveyance information including the layout and weight of the conveyance in the moving body, the weight of the moving body is obtained, the obtained first conveyance information and the weight of the moving body are input into a model generated using second conveyance information, the weight of the moving body, and the rudder angle center value of the moving body, thereby generating information representing a predicted rudder angle center value, where the second conveyance information is the past conveyance information with respect to the first conveyance information, the information representing the predicted rudder angle center value is output to the moving body.

2. The information processing method according to claim 1, the predicted rudder angle center value is set as the initial value of the rudder angle center when the moving body travels corresponding to the first conveyance information.

3. The information processing method according to claim 1 or 2, in the conveyance information, at least the layout of the conveyance is the reservation information of the moving body, the information representing the predicted rudder angle center value is generated after the reservation information is determined.

4. The information processing method according to claim 1 or 2, in the conveyance information, at least one of the layout and weight of the conveyance is generated using the output data of the weight sensor provided in the moving body.

5. The information processing method according to claim 1 or 2, in the conveyance information, at least one of the layout and weight of the conveyance is generated using the detection result of the conveyance, where the detection result of the conveyance is a detection result based on the output data of the camera sensor provided in the moving body.

6. The information processing method according to claim 5, the detection result includes the identification result of the conveyance, in the conveyance information, at least one of the layout and weight of the conveyance is generated using the attribute information of the identified conveyance.

7. The information processing method according to claim 1 or 2, the model is a model generated using a predefined basic model.

8. The information processing method according to claim 7, the model is a model generated using the center of gravity position of the moving body, the rudder angle center value, and the basic model, where the center of gravity position of the moving body is calculated based on the second conveyance information and the weight of the moving body.

9. The information processing method according to claim 1 or 2, the model is a model generated by performing machine learning with the second conveyance information and the weight of the moving body as training data and the rudder angle center value as correct answer data.

10. The information processing method according to claim 9, the model is a model generated using the training data further including the attribute information of the conveyance.

11. The information processing method according to claim 9, the model is a model generated using the training data further including the external environment information of the moving body.

12. An information processing system, comprising: A first acquisition unit that acquires first conveyance object information, which is conveyance object information including the layout and weight of a conveyance object in a mobile body; A second acquisition unit that acquires the weight of the mobile body; A generation unit that inputs the acquired first conveyance object information and the weight of the mobile body into a model generated using second conveyance object information, the weight of the mobile body, and a rudder angle center value of the mobile body, thereby generating information representing a predicted rudder angle center value, where the second conveyance object information is the conveyance object information in the past with respect to the first conveyance object information; And An output unit that outputs the information representing the predicted rudder angle center value to the mobile body.

13. A control device mounted on a mobile body, the control device comprising: A first acquisition unit that acquires first conveyance object information, which is conveyance object information including the layout and weight of a conveyance object in the mobile body; A second acquisition unit that acquires the weight of the mobile body; A generation unit that inputs the acquired first conveyance object information and the weight of the mobile body into a model generated using second conveyance object information, the weight of the mobile body, and a rudder angle center value of the mobile body, thereby generating information representing a predicted rudder angle center value, where the second conveyance object information is the conveyance object information in the past with respect to the first conveyance object information; And A setting unit that sets the predicted rudder angle center value as the rudder angle center value of the mobile body.

Images