JP2005184867A - Travelling speed pattern estimator, and drive controller of hybrid vehicle - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a travelling speed pattern estimator capable of estimating the travelling speed pattern of a travelling course easily. <P>SOLUTION: Travelling data and travelling environment data for a travelling course are stored while being associated with each other and a course travelled by a predetermined number of times or more is specified as a frequent course. The specified frequent course is divided into subsections at a point advanced by a predetermined distance (100 m) from a point where a start and a stop are predicted, e.g. an intersection, and a travelling speed pattern candidate is formed for each subsection. An estimated travelling speed pattern of a frequent course traveled from now is outputted by extracting a travelling speed pattern candidate matching current travelling environment data from each subsection. In addition to the intersection, the point where the start and the stop are predicted includes a point where a signal is installed, a tollgate, a curve start point, a junction, a lane decreasing point and the like. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a travel speed pattern estimation device and a hybrid vehicle drive control device, for example, estimation of a travel speed pattern for a frequently traveled route, and setting of an engine and motor operation schedule based on the estimated travel speed pattern. .

Hybrid vehicles using a combination of an engine such as an internal combustion engine and a motor such as an AC motor or a brushless DC motor rotated by electric power supplied from a power storage means such as a battery (secondary battery such as a storage battery) as a power source of the vehicle are widely used. It has been put into practical use.
Patent Document 1 proposes a technique for reducing engine combustion consumption in this hybrid vehicle.

JP 2000-333305 A

  In the technique described in Patent Document 1, a route to a destination is divided into a plurality of sections at points where start and stop are predicted, road data and a travel history are acquired from a navigation device, and travel for each section is performed. Based on the estimated traveling speed pattern and the fuel consumption characteristics of the engine, the engine and motor operation schedules are set so that the fuel consumption to the destination is minimized. .

In the technique described in Patent Document 1, if the start and stop prediction points for dividing a travel route are intersections, junctions, etc., the division points are clear and can be easily determined in connection with road information data. .
However, if the travel route is divided at the predicted start and stop points, there are the following problems from the viewpoint of travel prediction.
(1) Starting and stopping predicted points such as intersections are restraints and restrictive elements from the viewpoint of vehicle driving control, and are points where the driving state varies variously.
That is, an intersection or the like is a point where the vehicle passes as it is when it is a blue light, stops before it is red, and passes during the start. In addition, even if the vehicle starts, if there is a crossing pedestrian when turning left or right, the vehicle may stop and then restart. If there is a traffic jam, the stop position may leave the intersection. Furthermore, there are cases where the vehicle stops several times before the intersection and passes after starting.
(2) When predicting the traveling speed pattern, it is necessary to analyze the traveling data by dividing each section, but the traveling state at both ends of each section varies variously, and the analysis processing becomes complicated. For example, since passing, stopping, and starting are considered for each of the start point and end point of the divided section, it is necessary to consider the combination.
(3) Further, in the case of predicting traveling assuming that the analysis / classification has been completed, it is necessary to associate the predictions with the adjacent sections before and after, and it is not possible to analyze each section independently. For this reason, if a plurality of sections are considered in the travel route, the number of combinations becomes enormous.

Therefore, a first object of the present invention is to provide a travel speed pattern estimation device capable of easily estimating a travel speed pattern of a travel route.
It is a second object of the present invention to provide a drive control device for a hybrid vehicle that can easily set the operation schedule of the engine and the motor.

In the first aspect of the present invention, a travel data acquisition unit that acquires and stores vehicle travel data for a travel route, a frequent route identification unit that identifies a frequent route that travels frequently based on the travel data, Based on the travel data, the section dividing means for dividing the frequent route into small sections at a point traveled by a predetermined distance from the point where the start and stop are predicted, and similar travel data for the divided small sections A classifying means for classifying each data, and an estimated traveling speed pattern output means for generating a traveling speed pattern representative of the classification for each small section and outputting an estimated traveling speed pattern of a frequent route from which the vehicle travels. The first object is achieved by being provided in a velocity pattern estimation device.
According to a second aspect of the present invention, in the travel speed pattern estimation device according to the first aspect, the travel data acquisition means stores travel environment data in association with travel data for a travel route, and the estimated travel speed pattern. The output means extracts a traveling speed pattern candidate that matches the current traveling environment data for each of the small sections, and outputs an estimated traveling speed pattern of a frequent route to be traveled from now on.
According to a third aspect of the present invention, in the travel speed pattern estimation device according to the first or second aspect, the section dividing means includes a degree of traffic jam, a legal speed, a road width, and a travel area. The frequent route is divided into small sections at a point where the value of the predetermined distance is changed according to at least one element.
According to a fourth aspect of the present invention, an engine in which part or all of the driving force is used for driving or generating electric power of the vehicle and a motor for generating the driving force of the vehicle are provided, and driving of at least one of the engine and the motor A drive control device for a hybrid vehicle that travels by force, wherein the storage unit is charged by power generation and regeneration by driving the engine and supplies power to the motor, and a storage amount detection unit that detects a storage amount of the storage unit The travel speed pattern estimation device according to claim 1, claim 2, or claim 3, the estimated travel speed pattern of the frequent route output by the travel speed pattern estimation device, and the storage amount of the power storage means Based on the operation schedule setting method for setting the operation schedule of the engine and the motor so that the fuel consumption in the frequent route is minimized. When, by provided to the drive control device for a hybrid vehicle to achieve the second object.

According to the first to third aspects of the present invention, frequent routes are not performed at a point such as an intersection where start and stop are predicted, but are traveled at a predetermined distance from a point where start and stop are predicted. Is divided into small sections, so that the travel speed pattern of the travel route can be easily estimated.
Further, according to the present invention described in claim 4, since the estimated traveling speed pattern for the frequent route divided at a point traveled by a predetermined distance from the point where start and stop are predicted is used, the engine and motor operation schedules are used. Can be set easily.

Hereinafter, preferred embodiments of a travel speed pattern estimation device and a hybrid vehicle drive control device according to the present invention will be described in detail with reference to FIGS. 1 to 12.
(1) Outline of Embodiment In this embodiment, travel data and travel environment data for a travel route are stored in association with each other, and a route that has traveled a predetermined number of times or more is specified as a frequent route. The identified frequent route is divided into small sections at points that have traveled by a predetermined distance from points such as intersections where starting and stopping are predicted, and travel speed pattern candidates are created for each small section. Then, by extracting a traveling speed pattern candidate that matches the current traveling environment data for each small section, an estimated traveling speed pattern of a frequent route to be traveled is output.
In addition to intersections, there are points where traffic lights are installed, toll booths, start points of curves, merge points, lane decrease points, and so on.
And when dividing | segmenting into a small area in the point which advanced only the predetermined distance from the intersection etc., the fixed distance 100 [m] is selected as a predetermined distance.

The predetermined distance may be changed according to at least one element among the degree of traffic jam, legal speed, road width, and travel area. In other words, the greater the degree of traffic congestion, the higher the legal speed, and the wider the road width, the longer the predetermined distance. The smaller the degree of traffic congestion, the lower the legal speed, and the shorter the road width, the shorter the predetermined distance. Further, as a traveling area, the predetermined distance is increased in the urban area, and the predetermined distance is decreased in the suburbs.
As described above, in each small section divided at a point traveled by a predetermined distance from the intersection or the like without being divided at the intersection where the traveling state is unstable, the traveling state at both ends of the small section is stabilized. As a result, the conditions for both ends of the small section can be considered in a uniform state, and it is not necessary to consider the influence of the traveling state of the preceding and following sections, and only each small section can be analyzed independently.
A change in the driving state in the vicinity of the intermediate intersection in the small section can be grasped as a pattern, and classification becomes easy.
In addition, it is possible to independently analyze and predict independently in a plurality of continuous sections along the travel route, and connect the results without restriction.

  Further, in this embodiment, when classifying the transition pattern of the traveling speed in a certain small section, traveling such as an intersection is performed rather than the magnitude of the variation in traveling speed in a stable section where the traveling state is relatively stable each time. Classification is performed after weighting the travel data so that the magnitude of the change in travel speed in the variable section where the state is likely to change each time is large.

The traveling pattern predicting unit 11 (FIG. 1) accumulates traveling data on a frequent route on which the driver frequently travels, and divides the traveling data into small sections at division points set on the frequent route, Section travel data for each section is generated.
Each section travel data is weighted and analyzed as described above, and is classified into similar section travel data.
Further, the travel pattern predicting unit 11 generates and stores a typical travel speed pattern for the small section using the section travel data belonging to each classification. The generated traveling speed pattern may be single or plural.
In this way, the travel pattern prediction unit 11 prepares a typical travel speed pattern for each small section that constitutes the frequent route.

When the driver starts the vehicle, the traveling pattern predicting unit 11 predicts that the driver will travel on the frequent route from the current time and current position at that time.
Then, the travel speed pattern generated earlier is connected to each adjacent small section to generate a transition of the travel speed change of the entire frequent route, that is, an estimated travel speed pattern.
If there are multiple travel speed patterns in one small section, use the travel environment data at that time (weather, day of the week, etc.), the past travel frequency (frequency that traveled with the travel speed pattern), etc. Select a speed pattern.

The travel pattern prediction unit 11 generates an operation schedule of the engine 21 and the motor 24 using the estimated travel speed pattern generated in this way so that the fuel consumption is minimized.
Then, the travel pattern prediction unit 11 provides this operation schedule to the main control device 26 (FIG. 1), and the main control device 26 controls the engine 21 and the motor 24 using this schedule.
Thus, in this embodiment, when classifying the section travel data, in order to increase the data specific gravity in the variable section where the difference in travel state is most likely to appear, the section travel data can be classified more precisely, Thereby, a more appropriate traveling speed pattern can be extracted.

(2) Details of Embodiment FIG. 1 is a conceptual diagram showing a configuration of a drive control system for a hybrid vehicle to which a travel speed pattern estimation device and a drive control device for a hybrid vehicle according to an embodiment of the present invention are applied.
In FIG. 1, 10 is a hybrid vehicle drive control system as a travel speed pattern estimation apparatus in the present embodiment, and 20 is a drive apparatus. Here, reference numeral 21 denotes an engine such as an internal combustion engine driven by fuel such as gasoline or light oil, which includes an engine control device such as an ECU (not shown) and is used as a power source for vehicles such as passenger cars, buses and trucks. .

The driving force of the engine 21 is transmitted to a driving force transmission device 25 including a transmission (multi-stage transmission or continuously variable transmission) (not shown), a driving shaft, driving wheels, and the like, and the vehicle is rotated by rotating the driving wheels. Driven. The driving force transmission device 25 may be provided with a braking device such as a drum brake or a disc brake.
In the hybrid vehicle in the present embodiment, a part of the driving force of the engine 21 may be output for driving, and the remaining driving force may be used for driving the generator 22 to generate power. As a configuration for that purpose, for example, a planetary gear is used and the shafts of the engine 21, the generator 22, and the motor 24 are connected.

Here, the vehicle is a hybrid vehicle, has a motor 24 (AC motor, DC brushless motor, etc.) that is rotated by electric power, and uses the engine 21 and the motor 24 together as a power source for the vehicle. The motor 24 generates a driving force by the electric power supplied from the battery 23 as the power storage means, and the driving force is transmitted to the driving wheels of the driving force transmission device 25.
In addition, a generator 22 such as an AC generator is connected to the driving force transmission device 25 so as to generate a regenerative current when the vehicle is decelerated. The regenerative current generated by the generator 22 is supplied to the battery 23, and the battery 23 is charged.

  In addition, a current can be generated in the generator 22 by the driving force of the engine 21. The motor 24 is preferably an AC motor, and in this case, includes an inverter (not shown). Similarly, the generator 22 is preferably an AC generator, and in this case, includes an inverter (not shown). Further, the battery 23 includes a capacity detection sensor (not shown) for detecting SOC (State Of Charge) that is the amount of stored electricity. This capacity detection sensor functions as a storage amount detection means.

The motor 24 may be configured integrally with the generator 22. In this case, the motor 24 generates a driving force when electric power is supplied from the battery 23 and functions as a power source. When the motor 24 is rotated by the driving force transmission device 25 such as during braking of the vehicle, the regenerative current is generated. It functions as the generator 22 which generate | occur | produces.
Further, the generator 22 may generate power using the driving force of the engine 21, and the motor 24 may generate regenerative power during the deceleration operation of the vehicle.

The battery 23 is a secondary battery as a power storage means that can be repeatedly charged and discharged, and is generally a lead storage battery, a nickel cadmium battery, a nickel metal hydride battery, or the like, but is a high performance lead used in an electric vehicle or the like. A storage battery, a lithium ion battery, a sodium sulfur battery, etc. may be sufficient. Note that the power storage means does not necessarily have to be the battery 23, and electrically stores and discharges energy, such as a capacitor (capacitor) such as an electric double layer capacitor, a flywheel, a superconducting coil, a pressure accumulator, or the like. Any form may be used as long as it has a function.
Furthermore, any of these may be used alone, or a plurality of them may be used in combination. For example, the battery 23 and the electric double layer capacitor can be combined and used as a power storage means.

Reference numeral 26 denotes a main control device, which is a kind of computer including a calculation means (not shown) such as a CPU and an MPU, a storage means such as a semiconductor memory and a magnetic disk, a communication interface, and the like. And based on the signal from the various sensors of the driving | running | working data acquisition part 13, the operation | movement of the engine 21, the engine control apparatus, the motor 24, the generator 22, and an inverter is controlled.
Here, the sensor is an accelerator sensor, a brake sensor, or the like, and detects information related to the operation of the driver of the vehicle and transmits it to the main control device 26.

  The main control device 26 normally controls the usage ratio of the engine 21 and the motor 24 according to the traveling state of the vehicle and the storage state of the battery. For example, when the engine efficiency is low such as when starting or running at a low speed, the motor travels, and when the vehicle speed is higher than a predetermined vehicle speed or higher than a predetermined load, the engine travels. When the storage state of the battery is low, power may be generated in an efficient engine state and stored in the battery. When decelerating or downhill, regeneration is performed and the battery is charged.

In the present embodiment, the frequent route is divided into small sections at points that have traveled a predetermined distance from points where starting and stopping such as intersections and railroad crossings are predicted (hereinafter referred to as predicted starting points).
Then, the travel data for each divided small section is classified for each similar travel data, a travel speed pattern representing the classification is generated, and an estimated travel speed pattern of a frequent route to be traveled is determined. Based on the determined estimated traveling speed pattern, an operation schedule that minimizes the fuel consumption of the engine 21 is set for each small section.
When traveling on a frequent route, the engine 21 and the motor 24 are driven according to the set operation schedule.

In FIG. 1, 12 is a navigation database in which navigation information is stored as data used for navigation processing in a normal navigation device such as map data, road data, search data, and 14 is time, date and time, traffic jam information, It is a driving environment data acquisition part which acquires the data regarding the driving environment of vehicles, such as weather information. The travel data acquisition unit 13 includes various sensors, and acquires data related to the vehicle travel state such as the vehicle speed, the brake operation state, and the accelerator opening.
The travel pattern prediction unit 11 is a kind of computer that includes a calculation unit such as a CPU and MPU (not shown), a semiconductor memory (ROM, RAM, etc.), a storage unit such as a magnetic disk, a communication interface, and the like. Data is acquired from the travel data acquisition unit 13 and the travel environment data acquisition unit 14, and navigation processing such as display of the current position of the vehicle and route search to the destination is executed.
The travel pattern prediction unit 11 also reflects the driving characteristics of the driver, including a dividing point setting process that divides the traveling route at a point that has traveled a predetermined distance from the predicted stop / start point, and a frequent route dividing process. A running speed pattern estimation process for estimating the speed pattern is executed.
Furthermore, the travel pattern prediction unit 11 analyzes past travel data as data analysis for performing the travel speed pattern estimation process, and also performs a travel speed pattern extraction process in each small section.

  The travel pattern predicting unit 11 includes an input unit (not shown) including an operation key, a push button, a jog dial, a cross key, a remote controller, a CRT display, a liquid crystal display, an LED (Light Emitting Diode) display, a plasma display, and a windshield. Display unit including hologram device for projecting hologram, voice input unit including microphone, voice synthesizer, voice output unit including speaker, FM transmitter, telephone line network, Internet, mobile phone network, etc. It is desirable to have a communication unit that transmits / receives various data to / from.

  The navigation database 12 includes a database composed of various data files. In addition to search data for searching for a driving route, the navigation database 12 displays a guide map along the searched driving route on the screen of the display unit. A map to display characteristic photographs, frame diagrams, etc. at an intersection or travel route, to display the distance to the next intersection, the direction of travel at the next intersection, etc., or to display other guidance information Record various data such as data and facility data. The navigation database 12 also records various data for outputting predetermined information by the audio output unit.

  Here, the search data includes intersection data, road data, traffic regulation data, and route display data. In the intersection data, in addition to the number of intersections in which the data is stored, data relating to each intersection is assigned with a number for identification as intersection data. Further, each intersection data is stored with a number for identifying each connection road in addition to the number of roads connected to the corresponding intersection, that is, the number of connection roads. The intersection data may include the type of the intersection, that is, the distinction between the intersection where the traffic signal is installed or the intersection where the traffic signal is not installed.

  In addition to the number of roads in which the data is stored, the road data is stored with data relating to each road assigned with a number for identification as road data. Each road data stores a road type, a distance as a length of each road, a construction time as a time required for traveling on each road, and the like. Furthermore, the road type includes administrative road attributes such as national roads, prefectural roads, main local roads, general roads, and highways.

  In the road data, the width, gradient, cant, altitude, bank, road surface condition, whether there is a median, the number of road lanes, the point where the number of lanes decreases, the width of the road data is narrow. It is desirable to include data such as points. In the case of an expressway or a main road, each of the lanes in the opposite direction is stored as separate road data and processed as a double road. For example, in the case of a trunk road with two or more lanes on one side, it is processed as a two-way road, and the upward lane and the downward lane are stored in the road data as independent roads. Further, for corners, it is desirable to include data such as radii of curvature, intersections, T-junctions, and corner entrances. Road attributes such as railroad crossings, expressway entrance / exit rampways, expressway toll gates, downhill roads, and uphill roads may also be included.

Data stored in the navigation database 12 is stored in storage means such as a semiconductor memory or a magnetic disk. The storage means includes all forms of storage media such as magnetic tape, magnetic disk, magnetic drum, flash memory, CD-ROM, MD, DVD-ROM, optical disk, MO, IC card, optical card, memory card, etc. It is also possible to use a removable external storage medium.
The navigation database 12 may be placed on a server device operated by a navigation company, and road data may be provided to the travel pattern prediction unit 11 by wireless communication.
Since the state of the road changes frequently due to the opening of a new road, widening work, etc., the navigation data 12 can be easily updated / maintained by placing the navigation database 12 in the server device of the navigation company.

  The travel data acquisition unit 13 includes a GPS sensor that receives GPS information from a GPS (Global Positioning system) satellite, an azimuth sensor that detects the azimuth of the vehicle, an accelerator opening sensor that detects the accelerator opening, and driving. Brake switch for detecting the movement of the brake pedal operated by the driver, steering sensor for detecting the steering angle of the steering operated by the driver, the winker sensor for detecting the movement of the winker switch operated by the driver, and the shift operated by the driver A shift lever sensor that detects the movement of the machine's shift lever, a vehicle speed sensor that detects the vehicle speed, that is, a vehicle speed sensor, an acceleration sensor that detects the vehicle acceleration, and a yaw rate that indicates a change in the orientation of the vehicle. It has a yaw rate sensor. The travel data includes the current position of the vehicle, the accelerator opening, the movement of the brake pedal operated by the driver, the steering angle of the steering operated by the driver, the movement of the blinker switch operated by the driver, and the driver operating It includes the movement of the shift lever of the transmission, the vehicle speed, the acceleration of the vehicle, the yaw rate indicating changes in the direction in which the vehicle is facing, and the like.

Here, the travel data acquisition unit 13 acquires travel data such as the current position and travel speed of the vehicle at predetermined intervals from when the vehicle starts at the departure point to when it stops at the destination point. That is, travel data at a predetermined time interval (for example, every predetermined time such as 100 [msec], 1 [sec]) or a predetermined distance interval (for example, every predetermined distance such as 100 [m], 500 [m]). To get.
Or although it is an indefinite interval, you may acquire driving | running | working data corresponding to the point which can specify positions, such as every node point (delimitation point on the road database) of a road, or an intersection. In this way, by obtaining travel data at predetermined intervals, it is possible to obtain a trajectory of the travel of the vehicle and a change in travel speed at that time, that is, a travel speed pattern. Can be used to generate and estimate

  The travel environment data acquisition unit 14 includes a clock, a calendar, and the like, and acquires and stores date and time information such as the current time, date, day of the week, and date and time the vehicle departs. In addition, the traveling environment data acquisition unit 14 collects information on traffic control systems such as police and the Japan Highway Public Corporation in a road traffic information communication system called VICS® (Vehicle Information & Communication System), for example. Road traffic information such as information on traffic jams, traffic regulation information, construction information on road construction, etc. is acquired and stored.

Furthermore, the driving environment data acquisition unit 14 specifies event information such as scheduled locations and scheduled dates of events such as festivals, parades, fireworks displays, etc., for example, roads around stations and large commercial facilities every day except weekends. It is desirable to obtain and store statistical traffic information such as traffic jams occurring at the time of day, traffic jams occurring during summer holidays, and weather information such as weather forecasts created by the Japan Meteorological Agency. .
The travel environment information as information on the environment in which the vehicle travels acquired and stored by the travel environment data acquisition unit 14 includes the current time, date, day of the week, date and time of departure of the vehicle, weather, traffic congestion information, and traffic regulation information. , Road construction information, event information, etc.

The travel environment data acquisition unit 14 also acquires the date and time, day of the week, wiper, headlight, air conditioner, operation status of vehicle-mounted devices such as a defroster, and sensing data of vehicle-mounted sensors such as a raindrop sensor and a temperature sensor. To do.
The operation status data and sensing data of the on-vehicle equipment can be used by the travel pattern prediction unit 11 to estimate the weather at that time.

The travel data storage unit 15 stores travel data acquired by the travel data acquisition unit 13 and travel environment data acquired by the travel environment data acquisition unit 14.
In this case, the travel data and the travel environment data in one travel of the vehicle are stored in association with each other. The travel speed pattern candidate storage unit 16 stores a travel speed pattern generated from past travel data described later.

Further, the division point storage unit 17 stores the data of each division point set in the division point setting process for each travel route. This division point data is stored for each travel route specifying data for specifying a travel route (a data string composed of departure point, intersection existing on the travel route, and each point data specifying the arrival point).
The dividing point setting process in the present embodiment is executed for each travel route after the initial travel based on the travel data acquired when traveling on the travel route for the first time. In addition, it may be executed after a predetermined number of runs.

In the drive control system 10 for the hybrid vehicle of the present embodiment, the traveling pattern prediction unit 11 has a function as a traveling speed data classification device and a function as a traveling speed pattern estimation device.
When the travel pattern prediction unit 11 functions as a travel speed data classification device, the travel pattern prediction unit 11 includes a travel data acquisition unit, a travel route identification unit, a frequent route identification unit, a section division unit, a weighting unit, a classification unit, and an estimated travel. It functions as a speed pattern output unit and a weight release unit.

Among these means, the travel data acquisition means acquires the travel data of the vehicle and stores it in the travel data storage unit 15. In addition, the travel data acquisition unit uses the travel data already stored in the travel data storage unit 15 and the information on the division points stored in the division point storage unit 17, so that the travel data in the small section to be analyzed (Hereinafter, section travel data) is acquired.
The travel data acquisition means acquires travel environment data of the vehicle such as weather during travel, time, date, day of the week, etc., when a travel pattern prediction unit 11 described later functions as a travel speed pattern estimation device.

The travel route specifying means specifies an arbitrary travel route.
For example, the travel route specifying means predicts and specifies a travel route from which the user will travel. This prediction can be performed, for example, by determining that there is a high probability that the driver always uses a road (frequent route) from the current position and current time at the start of the drive control system 10 of the hybrid vehicle. Is possible.

The frequent route specifying means specifies a frequent route that frequently travels based on the travel data.
That is, the frequent route specifying means specifies, for example, a travel route frequently used repeatedly by the driver, such as a commuting route, as the frequent route. More specifically, for each travel route that has traveled in the past, travel data and travel environment data are stored in the travel data storage unit 15 in association with each other. Among them, a travel route that has traveled a predetermined number of times or more is identified as a frequent route.

The section dividing means divides the travel route specified by the travel route specifying means into small sections.
When dividing into small sections, the section dividing means uses the dividing point data set in the dividing point setting process after the first run and stored in the dividing point storage unit 17.

  The weighting means multiplies the past travel data of the small section by a weighting coefficient corresponding to the position in the small section. In this case, the weighting means sets the weighting coefficient of a predetermined section including a point where start and stop are predicted in a small section to a value larger than the weighting coefficient other than the predetermined section. As a result, travel data in which variations in travel speed change are increased is generated.

  Here, “weighting according to the travel position in the small section” means a position where the vehicle such as an intersection, a temporary stop, a toll gate, etc. is estimated to stop in the small section (these are determined from the navigation map data). ), For example, a section within 100 [m] before and after the position estimated to stop is multiplied by a weighting factor larger than the weighting coefficient applied to the traveling data of the section other than the section. It is.

The classifying unit classifies the traveling data having a large change in the traveling speed by the weighting unit, for example, for each traveling data that is similar in the transition of the variation in the traveling speed.
Here, as a method of “classifying according to the similarity of the transition of the travel speed change” by the classification means, a clustering method called a k-average method, a method for obtaining the similarity of time series graph curves of travel data, For example, a method of classifying a group of travel data in which variation in travel data for each position in the section falls within a predetermined range as a group.

  For example, when classifying the data where the variation in travel data for each position in a small section falls within a predetermined range as one group, the above-mentioned weighting expands the variation, so it can be classified into one group without weighting. Even things may be classified into a plurality of groups. In other words, the weighting brings about an effect that the difference in the running data of the weighted section is reflected in the classification, and the classification accuracy can be improved.

The estimated traveling speed pattern output means generates a traveling speed pattern representing the classification for each classification of traveling data similar to the divided subsections by the classification means and stores it in the traveling speed pattern candidate storage unit 16. To do.
The generation of the small section speed pattern can be performed, for example, by averaging the traveling data of each classified small section.
The estimated traveling speed pattern output means associates the traveling environment data of the traveling data contributing to the generation of each traveling speed pattern with the traveling speed pattern when the traveling pattern predicting unit 11 described later functions as a traveling speed pattern estimating device. And stored in the traveling speed pattern candidate storage unit 16.

  The weighting cancellation unit cancels the weighting of the travel data classified by the classification unit and restores the original travel data of the small section. Thereby, section travel data grouped (aggregated) for each classification is obtained.

Further, when the travel pattern prediction unit 11 functions as a travel speed pattern estimation device, the travel pattern prediction unit 11 functions as an estimated travel speed pattern output unit.
The estimated travel speed pattern output means extracts a travel speed pattern that matches the travel environment data of the vehicle when leaving an arbitrary travel route to be traveled for each small section, and estimates the entire travel route to be traveled from now on Output travel speed pattern.
That is, the estimated traveling speed pattern output unit acquires traveling speed pattern candidates from the traveling speed pattern candidate storage unit 16. When there are a plurality of travel speed patterns in one small section, the travel environment data and past frequency are referred to, and the one that is predicted to be most likely to travel on that day is selected.
Then, the estimated travel speed pattern output means connects the selected travel speed pattern for each adjacent small section with respect to the entire travel route specified by the travel route specifying means, and outputs an estimated travel speed pattern over the entire travel route. Output.

The main control device 26 functions as an operation schedule setting unit, and based on the estimated traveling speed pattern output by the estimated traveling speed pattern output unit and the amount of power stored in the power storage unit, the fuel consumption amount on the frequent route is minimized. In addition, the operation schedule of the engine 21 and the motor 24 is set.
More specifically, this operation schedule is performed by planning charging / discharging of the battery 23 within the SOC management range.

As described above, the driving pattern prediction unit 11 is (a) a function for analyzing a driving speed pattern using past driving data, and (b) a driver will drive based on the analyzed information. It has the function to predict a travel speed pattern with respect to the travel route. Hereinafter, these functions will be described in detail.
Note that the main control device 26 sets the operation schedule based on the predicted travel speed pattern.

(A) Analysis Function for Estimating Travel Speed Pattern In this embodiment, as an example, an analysis target of travel data and an operation schedule setting target are frequent routes. Here, the frequent route is, for example, a travel route that is frequently used repeatedly by the driver, such as a commuting route.
More specifically, the vehicle according to the present embodiment stores travel data and travel environment data for the travel route in association with each other in the travel data storage unit 15, and the travel pattern prediction unit 11 stores the travel route data. Among them, a travel route that has traveled a predetermined number of times or more is identified as a frequent route.
In addition, this does not limit the analysis target to the frequent route, but can be the analysis target as long as the travel route stores past travel data.

Regarding frequent routes, it is considered that a pattern exists in the running state because the same route is repeatedly traveled, and it is possible to estimate this travel speed pattern from past travel data.
Therefore, when the driver intends to travel on a frequent route from now on, the traveling speed pattern is estimated in advance, and the operation schedule of the engine 21 and the motor 24 is set so that the fuel consumption is minimized based on the estimation. Is possible.

FIG. 2 is a diagram for explaining an example of a method of dividing a frequent route into small sections.
In order to analyze the travel speed pattern for the entire frequent route, the travel pattern prediction unit 11 divides the frequent route into small sections and analyzes the travel speed pattern for each of the small sections. The analyzed traveling speed pattern for each small section is connected to each adjacent small section to generate a traveling speed pattern for the entire frequent route.
Here, first, a method for dividing a frequent route will be described with reference to FIG. Note that FIG. 2 is a diagram schematically showing a frequent route in order to explain the dividing method.

In the frequent route 30, there are stop and start predicted points 33 a to 33 c where start and stop are predicted from the starting point 31 to the destination 32.
Here, the predicted stop / start point where start and stop are predicted is, for example, after the vehicle has decelerated or stopped, such as an intersection, other traffic signal installation locations, railroad crossings, T-junctions, temporary stop line installation locations, etc. This is the place where it is predicted to start and accelerate.

The traveling pattern prediction unit 11 divides a point located at a predetermined distance in the traveling direction of the vehicle from the predicted stop starting point, that is, a point moved (slid) by a predetermined distance in the predicted traveling start point in the traveling direction of the vehicle. The points are identified as points, and the frequent route is divided at these dividing points.
As a result, the frequent route 30 is divided at the dividing points 35a to 35c, and as a result, the frequent route 30 is divided into four small sections 1 to 4 from the departure point to the destination.

  As described above, the traveling pattern prediction unit 11 divides the frequent route 30 with the stop / start predicted points 33a to 33c as a reference, but does not divide the stop / start predicted points 33a to 33c, but from the stop / start predicted points 33a to 33c. It divides | segments by the dividing points 35a-35c in the position slid by predetermined distance.

In the division point data stored in the division point storage unit 17, information for specifying the division points 35a to 35c is recorded by the division point setting process described below.
That is, when the vehicle starts to travel, the travel pattern prediction unit 11 acquires travel data from engine start to stop together with travel environment data and temporarily stores the travel data.
Then, the travel pattern prediction unit 11 determines whether or not the travel route from which the travel data has been acquired is a route that has traveled in the past (whether or not there is travel data for the same travel route). For example, it is stored in the travel data storage unit 17 together with the travel environment data as travel data for the travel route.

On the other hand, in the case of the first travel, the travel pattern prediction unit 11 sets the departure point PD0 as the division start point P0, detects the predicted stop / start point on the travel route, and sets Pn for each.
Then, the traveling pattern prediction unit 11 sets a point obtained by adding a predetermined distance in the vehicle traveling direction to each stop / start predicted point as a division point PDn (PDn = Pn + predetermined distance). Further, the traveling pattern prediction unit 11 sets the destination as the division end point (PDn + 1 = Pn + 1).
As described above, the division points (including the division start point and the division end point) set for the initial travel route are stored in the division point storage unit 17 as division point data.

In this way, the traveling pattern predicting unit 11 does not divide the frequent route at the stop / start predicted point where the traveling state is indeterminate, and further divides at the point that has traveled by a predetermined distance. The running state is stable.
As a result, the conditions at both ends of the small section can be considered, and there is no need to consider the influence of the traveling state of the preceding and following sections, and only each small section can be analyzed independently.
In addition, since the traveling state of the vehicle at both ends of the small section is constant, it is easy to connect the traveling speed pattern in the adjacent small section. Further, the traveling state change near the intermediate intersection in the small section is a pattern. It can be grasped and classification becomes easy.
In addition, it is possible to independently analyze and predict independently in a plurality of continuous sections along the travel route, and connect the results without restriction.

In the above dividing point setting process, the dividing point is not divided at a predicted stop / start point such as an intersection, but a certain distance (which can be a certain time or a certain time distance) is slid in the traveling direction of the vehicle. By dividing at the (advanced) point, classification and joining for each section become easy.
However, if the dividing point is too far from the predicted stop / start point, there is a problem that the processing becomes difficult in the following cases.

(A) A case where re-prediction is performed because the predicted traveling pattern (estimated traveling speed pattern) and the actual traveling are greatly different.
That is, it is assumed that the road and the road type change and the traffic flow changes at the predicted stop / start point, so that the predicted travel pattern and the actual travel tend to be different.
(B) When the travel route is changed.
That is, the actual travel route is changed from the planned travel route mostly at the predicted stop / start point such as an intersection.

  Therefore, when the distance from the predicted stop / start point to the split point (predetermined distance to be added to the predicted stop / start point) is large in the section dividing process, in the case of (a) and (b) above, it is in the middle of the divided section. In order to perform re-prediction and processing from that section, there are many parts that have finished traveling, which is inefficient. For this reason, the process of (1) and (2) can respond efficiently by dividing | segmenting in the point near a stop start prediction point.

Therefore, in the present embodiment, when the analysis is performed, the divided section is moved from the stop prediction predicted point by a fixed distance, a fixed time distance, or a fixed time (a point where the travel is relatively stable, and the stop start predicted point is as much as possible). Set the nearest point as the dividing point.
Specifically, as the predetermined distance in the present embodiment, 100 [m] is set as the distance that the traveling state of the vehicle is considered to be stable, but other distances may be used, for example, 50 [m] to 100 [m]. May be a distance n [m] between or a distance between them.

In the embodiment described above, the predetermined distance is fixed for each small section in order to facilitate the section division process, but the present invention is not limited to this.
That is, the distance from the predicted stop / start point to the relatively stable travel varies depending on the characteristics of the section and adjacent sections, such as distance, area, road surface, road width, legal speed, traffic flow, especially in traffic jams, etc. Therefore, for example, it can be set according to the size of the road (national road, prefectural road, city road, etc.), or an arbitrary distance can be set for each road according to the traffic volume or traffic condition. It is also possible to set.
In this case, the traveling pattern prediction unit 11 sets a predetermined distance (a certain time distance or a certain time) to be shifted from the predicted stop / start point according to the following examples (a) to (d).

(A) Adjust according to the degree of traffic jam.
When the vehicle is congested, it takes time for the traveling to become relatively stable. Therefore, a certain distance (a certain time distance or a certain time, hereinafter the same) is divided.
That is, the predetermined distance when there is a traffic jam is set to be longer than that when there is no traffic jam.
(B) Adjust according to the legal speed.
The time and distance until the driving becomes relatively stable for each legal speed is different. For example, a highway has a high statutory speed, so it takes a long time and a long distance to reach a relatively stable driving. Therefore, when the legal speed is high, the predetermined distance is made longer than when the legal speed is low.
In other words, on the municipal road, the legal speed is low, and the time until it becomes stable is shortened, so the predetermined distance may be short.

(C) Adjust according to the width of the road.
When the width of the road is narrow, it is difficult for the driver to feel that there are buildings on the side of the road and speed is difficult. For this reason, the time and distance until a stable traveling state in the section is shortened. Therefore, the predetermined distance is increased when the road is wide, and is shortened when the road is narrow.
(D) Adjust according to the region.
For example, in a city area, it takes a long time to reach a relatively stable state with a relatively heavy traffic, so the predetermined distance is increased. On the other hand, in mountainous areas and suburbs, since the traffic volume is small, the distance and time required for traveling to be relatively stable are short, so the predetermined distance is shortened.

The traveling pattern prediction unit 11 generates a traveling speed pattern in each small section by analyzing a plurality of past traveling data in the small section divided in this way.
Of the travel data from the departure point of the frequent route to the destination, the portion belonging to the small section being analyzed is referred to as section travel data.
The section travel data can be generated by extracting a portion corresponding to the small section being analyzed from the travel data.
Below, the classification | category method of a segment travel data and the production | generation method of a travel speed pattern are demonstrated using FIGS.

FIG. 3 is a diagram for explaining the weighting set for the section travel data.
The small section may include, for example, a section where the traveling state is approximately stable every time, such as a straight road, and a section where the traveling state changes depending on the time, such as near an intersection.
In the present embodiment, the former section where the traveling state is approximately stable is called a stable section, and the latter section where the traveling state changes is called a variable section.

The characteristic part that is a clue to typifying the transition of travel speed appears mainly in the variable section, so when classifying and patterning the section travel data, it is particularly important for the transition of the travel speed within the variable section. It is desirable to pay attention.
Therefore, the travel pattern predicting unit 11 reduces the fluctuation of the travel speed in the stable section to clearly recognize the transition of the travel speed change in the variable section, and changes the travel speed in the variable section. The travel data is weighted according to the position in the small section so as to increase.

In this embodiment, as shown in the weight distribution 40 (FIG. 3), for example, the data specific gravity in the stable section is set to 0.7, and the weighting is performed so that the data specific gravity in the fluctuation section is 1.0. . Here, the fluctuation section was set to a range of 100 [m] before and after the predicted stoppage.
Among the stable sections, the area in the range of 100 [m] from the boundary with the fluctuation section was set so that the data specific gravity gradually increased with a predetermined gradient from 0.7 to 1.0 toward the boundary. .
Note that the data specific gravity of the stable section may be 1.0, and the data specific gravity of the fluctuation section may be n (> 1.0). Also in this case, regarding the region within the range of 100 [m] from the boundary with the fluctuation interval in the stable interval, the data specific gravity is set so as to gradually increase with a predetermined gradient from 1.0 to n toward the boundary. May be.

The travel pattern prediction unit 11 performs weighting by multiplying 0.7 by the travel speed in the stable section and multiplying the travel speed in the variable section by 1.0 according to the weight distribution 40. However, for the region of 100 [m] from the boundary of the variable section in the stable section, the traveling speed is multiplied by a value from 0.7 to 1.0 according to the weight distribution 40.
Hereinafter, the weighted section travel data is referred to as weighted section travel data.

Thus, by making the data specific gravity in the fluctuation section larger than the data specific gravity in the stable section, it is possible to make the difference in travel speed fluctuation in the fluctuation section larger than the difference in travel speed in the stable section.
That is, in the fluctuation section, it is possible to further vary the value that varies, compared to the stable section, and to emphasize the characteristics of the speed change. Therefore, the weighted section travel data can be classified more clearly and easily than the section travel data.
In this way, by classifying the weighted section travel data, it is possible to classify the section travel data with more emphasis on the change in travel speed near the dividing point than when classifying the section travel data. A precise travel speed pattern can be generated.

Next, using FIG. 4, the section traveling data and the weighted section traveling data weighted according to the weight distribution 40 are shown. In each figure, the travel speed (unit [km / h]) is taken in the vertical axis direction and the distance (unit [m]) is taken in the horizontal axis direction, and the travel data for two runs are shown by a solid line and a dotted line, respectively.
FIG. 4A shows a part of traveling data on a certain traveling path. This travel path has intersections A, B, C,.
A small section 53 is formed by using a point slid in the traveling direction of the vehicle by a predetermined distance 51 from each intersection as a dividing point.
From FIG. 4A, it can be seen that the shape and position of the solid line and the dotted line are substantially the same in the region where the vehicle traveling state is stable, and the dividing point is the traveling stable point.

FIG. 4B shows weighted section travel data in which the small section 53 is divided into a stable section 57 and a variable section 58 and weighted according to the weight distribution 40.
As shown in FIG. 4B, the travel data 55 belonging to the stable section 57 corresponds to the travel data shown in FIG. 4A, whereas the travel speed is about 35 [km / h]. The traveling speed of the part is about 50 [km / h]. Thus, in the stable section 55, the traveling speed is 0.7 times.
On the other hand, the travel data 56 belonging to the fluctuation section 58 is 1.0 times that of the corresponding portion of the travel data shown in FIG.
For this reason, the variation in travel speed in the stable zone 57 is reduced to 0.7 times the variation in the fluctuation zone 58.
In this way, by weighting the travel data according to the weight distribution 40, the variation in travel speed in the variable section can be made relatively larger than the variation in travel speed in the stable section, and the difference in travel state can be reduced. It can be made clearer.

The traveling pattern predicting unit 11 statistically analyzes the traveling state in each small section from the plurality of past traveling data, and extracts a traveling speed pattern in each small section.
Hereinafter, a method of extracting and generating a traveling speed pattern from a plurality of traveling data will be described.
FIG. 5A is a diagram in which section traveling data of a certain small section is graphed, and FIG. 5B is a diagram in which the overlapping section traveling data generated from these section traveling data is graphed. In any case, a plurality of section travel data are overlaid.
In each graph, the vertical axis represents the vehicle traveling speed (unit [km / h]), and the horizontal axis represents the distance (unit [m]) from the starting point of the small section. A point where the vehicle enters a small section (departure side) is a start point, and a point where the vehicle exits (a destination side) is an end point.
By analyzing the plurality of section travel data, a travel speed pattern in this small section is generated.

In the graph of FIG. 5A, there is a predicted stop / start point at a point of about 600 [m] from the start point. An area of about 500 [m] from the start point is a stable section, and an area from the point of about 500 [m] to the end point is a variable section.
As shown in FIG. 5A, in the stable section, the traveling speed value is generally in the range of 30 to 60 [km / h], whereas in the fluctuation section, the traveling speed value is 0. It largely changes in the range of ˜60 [km / h].
In addition, there are various patterns in the variable section, such as a section that continuously passes from the stable section at a constant traveling speed and a section that accelerates after decelerating before the predicted stoppage.

As shown in FIG. 5B, when the section speed data is weighted according to the weight distribution 40, the transition of the change in the traveling speed in the variable section appears larger than that in the stable section.
As shown in FIG. 5 (b), in the stable section, the value of the traveling speed is generally within the range of 30 to 40 [km / h], the fluctuation range of the traveling speed becomes smaller, and the average speed is increased. Is smaller by 30 [%] than before weighting.
On the other hand, the transition of the change in travel speed in the variable section is the same as that before weighting.
When connecting the travel speed pattern in the small section later and setting the driving schedule for the entire travel path, the difference in the travel speed pattern in the variable section is important, so as in this embodiment, the travel in the variable section A method of classifying the section traveling data by paying attention to the speed pattern is very effective.

Next, a method for classifying weighted weighted section travel data according to a change in travel speed will be described.
As a method of classifying traveling data, for example, there are a case where classification is based on average speed and a case where classification is similar to a curve (hereinafter referred to as a traveling data curve). However, as described below, any method is used. However, the classification using the weighted section traveling data can be classified with higher accuracy than the case where the unweighted section traveling data is used.
In addition, by using a clustering technique called a k-average method, classification is made similar to each other, and an average traveling speed pattern is extracted for each classification.

In the classification using the average speed, the running speed is averaged over the entire target section, and the running data is classified based on the average value.
In the weighted section travel data, the variation in average speed in the stable section is reduced, so if the average value of the travel speed is taken over the entire small section, the contribution to the average value of the travel speed in the stable section Get smaller.
That is, the traveling speed in the fluctuation section is more largely reflected in the average value. Therefore, when the weighted section travel data is classified based on the average value, it is possible to perform classification that more accurately reflects the difference in travel speed pattern in the variable section.

On the other hand, as shown in FIG. 5B, the classification using the similarity of the travel data curve is the similarity of the travel data curve in the two-dimensional space in which the transition of the travel speed is expressed as a function of the distance from the starting point of the small section. The travel data is classified according to the above.
In the weighted section travel data, the difference in the travel data curve in the stable section is relatively smaller than the difference in the fluctuation section, so the difference in the travel data curve in the stable section is considered for the entire small section The influence on the difference in the similarity of the travel data curves is reduced.
Therefore, it is possible to perform classification that more accurately reflects the difference in travel speed pattern in the variable section.

FIG. 6 is a diagram in which the travel pattern prediction unit 11 classifies the weighted section travel data shown in FIG. 5B by using the similarity of the travel data curve. In the case of FIG.5 (b), as shown to Fig.6 (a)-(c), it can classify | categorize into the three patterns of the classification | category 1-classification | category 3. In each figure, a plurality of classified heavy section traveling data are displayed in an overlapping manner.
The following characteristics can be found from each classification.

In the classification 1 shown in FIG. 6A, the traveling speed in the stable section is constant between about 30 and 40 [km / h], and immediately before the predicted start / stop point (about 600 [m]). The vehicle is decelerating substantially to the stop state.
After deceleration, the vehicle accelerates again and moves to the next small section at a point (end point) where the traveling speed is substantially constant.

In the classification 2 shown in FIG. 6 (b), the traveling speed in the stable section is constant between about 30 and 40 [km / h], and the traveling speed at the predicted start / stop point (about 600 [m]). Is constant between 40 and 50 [km / h].
Considering that the traveling speed in the stable section is weighted by 0.7 times, it can be said that in classification 2, the traveling speed is almost constant over the entire small section from the start point to the end point.

In the classification 3 shown in FIG. 6C, the traveling speed in the stable section varies in the vicinity of about 30 [km / h], and decelerates before reaching the fluctuation section. In the fluctuating section, acceleration / deceleration is repeated, and the running speed varies even after passing through the predicted stop / start point.
As described above, when looking at the driving state in the variable section of each category, in category 1, the vehicle is accelerated after deceleration, in category 2, it is a constant speed, and in category 3, it is repeatedly accelerated and decelerated at a low speed. The difference in the weighted section travel data appears remarkably.
On the other hand, when looking at the traveling state in the stable section of each category, the traveling speed is within the same range in Category 1 and Category 2, and in Category 3, the fluctuation range is Category 1 Although it is wider than 2, it is still within a certain range.
From the above facts, it can be seen that by weighting the section travel data with the weight distribution 40, the difference in the variable section is emphasized and more accurate classification can be performed.

FIGS. 7A to 7C are graphs showing section travel data corresponding to the weighted section travel data after classification, and correspond to classifications 1 to 3 in FIG. 6, respectively.
Each section travel data after classification may be generated by canceling the weighting of the classified section travel data after classification, or the section travel data and the weighted section travel data are associated with each other on a one-to-one basis. The section traveling data may be classified in correspondence with the classified section traveling data after classification.
By generating a typical travel data curve for each category from each segment travel data belonging to these classifications 1 to 3 (extracting a representative pattern), a travel speed pattern representing each category can be generated. it can.

The following features and trends can be read from each classification.
In classification 1 shown in FIG. 7A, the traveling speed in the stable section is constant between about 50 and 60 [km / h]. And after decelerating in a section of about 100 [m] before the predicted stop / start point (about 600 [m]), the vehicle accelerates, passes through the predicted stop / start point, and again reaches about 50-60 [km / h]. The traveling speed is recovered and the vehicle moves to the next small section.
This pattern corresponds to a case where the traffic in the stable section is smooth and stopped or decelerated at the predicted stop / start point.
This may be the case, for example, when the vehicle is running but the vehicle starts to stop when it encounters a red light at an intersection.

In the classification 2 shown in FIG. 7B, the traveling speed in the stable section and the variable section is constant between about 40 and 60 [km / h].
In the case of classification 2, since no change in the traveling speed at the predicted stop / start point is observed, the traffic is smooth over the entire small section, and the stop / start predicted point passes smoothly.
For example, this may be the case when the road is running and the intersection can be passed with a green light.

In the classification 3 shown in FIG. 7C, the traveling in the stable section fluctuates between about 30 to 50 [km / h], and the traveling speed fluctuates as the moving section approaches.
Then, deceleration and acceleration are repeated from before reaching the fluctuating section to the stop / start predicted point, and after passing through the stop / start predicted point, the running fluctuates between about 30 to 50 [km / h]. From this, it can be seen that classification 3 corresponds to a case where traffic on the road is stagnant.
For example, there may be a case where the traveling road is congested (especially in the vicinity of the intersection), and after passing the intersection several times after waiting for a signal.

As described above, when segment travel data is analyzed and similar ones are classified, it can be seen that the segment travel data shows similar characteristics in each classification. Therefore, it is possible to extract a pattern representing each classification, that is, a traveling speed pattern, by statistically processing the classified section traveling data.
In this way, the travel pattern prediction unit 11 generates a representative travel speed pattern from the section travel data belonging to each classification.

FIGS. 8A to 8C are traveling speed patterns corresponding to the classifications 1 to 3 shown in FIG. 7, respectively, and are generated by the traveling pattern prediction unit 11 using the section traveling data belonging to the respective classifications. It is a thing.
In these travel speed patterns, the characteristics of the section travel data belonging to each category are extracted, and the travel speed patterns are representative of each category.
Various methods for generating the traveling speed pattern from the section traveling data classified into each classification are conceivable. In the present embodiment, the section traveling data is generated by averaging the section traveling data for each classification.

As another method of generating the traveling speed pattern, for example, the average of the section traveling data after excluding the most distant from the overall tendency (that is, excluding the traveling data curve having the most different shape) It can also be configured to.
First of all, the average of all section travel data is taken, and the difference between each section travel data is compared. And the thing with the largest difference is specified, and after removing this, the section traveling data is averaged again.
In this way, by averaging after removing the section travel data that is farthest from the overall trend, it is possible to reduce the variation of the section travel data to be averaged, and to more accurately extract the characteristics of the classification it can.

The traveling speed pattern shown in FIG. 8A is stable when the traveling speed in the stable section is about 53 [km / h]. In the varying section, the traveling speed pattern accelerates after decelerating before the predicted stoppage. ing.
Thus, this travel speed pattern reflects the characteristics of the section travel data shown in FIG. 7A and is a typical travel speed pattern of the section travel data to which the category 1 belongs.

In the traveling speed pattern of classification 2 shown in FIG. 8B, the traveling speed is constant at about 47 [km / h] over the entire small section.
Moreover, the traveling speed pattern of the classification 3 shown in FIG.8 (c) fluctuates between 40-50 [km / h] in the stable area, and decelerates before reaching a fluctuation area. . The acceleration / deceleration is repeated in the fluctuation section.
Both the traveling speed patterns of the classification 2 and the classification 3 reflect the characteristics of the section traveling data of the classification 2 and the classification 3 shown in FIGS. 7B and 7C, respectively.

Next, a procedure for generating a travel speed pattern from travel data will be described using the flowchart of FIG.
First, the traveling pattern prediction unit 11 specifies a frequent route to be analyzed (step 5).
This process is performed, for example, by acquiring past travel data from the travel data storage unit 15 and specifying a travel route that has traveled a predetermined number of times as a frequent route.
In addition, a frequently used traveling route such as a commuting route may be registered in advance as a frequent route by the driver, and the frequent route may be specified thereby.

Next, the traveling pattern prediction unit 11 extracts a plurality of pieces of past traveling data stored in the traveling data storage unit 15 relating to the frequent route identified in Step 5 (Step 10).
Next, the traveling pattern predicting unit 11 acquires the dividing point data related to the frequent route from the dividing point storage unit 17, and uses this to divide each traveling data acquired in step 10 into section traveling data for each small section. (Step 15). Thereby, a plurality of section travel data is obtained for each small section.
Moreover, the traveling pattern prediction unit 11 counts and stores the number of small sections (N) generated from the frequent routes.

Next, the traveling pattern prediction unit 11 sets the initial value of the counter n for counting small sections to 1 (step 20).
Then, the travel pattern prediction unit 11 weights each section travel data belonging to the section for the nth small section from the start point side of the frequent route with a predetermined distribution (for example, weight distribution 40), and performs the weighted section travel. Data is generated (step 25).
Next, the traveling pattern predicting unit 11 classifies the weighted section traveling data according to a predetermined method such as the similarity of traveling data curves (step 30).

When the traveling pattern prediction unit 11 finishes classifying the weighted section traveling data, the traveling pattern prediction unit 11 releases the weighting of the weighted section traveling data and restores the section traveling data (step 35). Thereby, section travel data classified for each classification is obtained.
Next, the traveling pattern prediction unit 11 generates a traveling speed pattern for each classification (step 45). Thereby, a typical traveling speed pattern representing each classification can be obtained.
Then, the travel pattern prediction unit 11 stores the generated travel speed pattern in the travel speed pattern candidate storage unit 16 in association with the subsection to be analyzed and the classification to which the travel speed pattern belongs.

The traveling pattern prediction unit 11 ends the analysis for the nth small section through the above steps 25 to 45. Then, n and N are compared, and it is determined whether or not analysis has been performed for all the small sections (step 50).
If n is smaller than N (step 50; Y), there is a small section that has not been analyzed yet, so 1 is added to n (step 55), and the process returns to step 25.
On the other hand, if n = N (step 50; N), the analysis is completed for all the small sections, and the process is terminated.

Through the above processing, the travel speed pattern candidate storage unit 16 can store the travel speed patterns classified by the pattern over all the small sections constituting the frequent route.
As described above, the analysis is performed with respect to a certain frequent route. When a plurality of frequent routes are identified, the traveling pattern prediction unit 11 generates a traveling speed pattern by analyzing each frequent route and stores it in the traveling data storage unit 15. To do.

In the present embodiment, the travel speed pattern is generated in advance and stored in the travel speed pattern candidate storage unit 16. However, the present invention is not limited to this, and each time the vehicle travels on the frequent route, It can also be configured to be generated and used every time.
Alternatively, the travel speed pattern may be generated again each time new travel data relating to the frequent route is obtained, and the travel speed pattern stored in the travel speed pattern candidate storage unit 16 may be updated.
At that time, it is configured so that newer driving data is reflected in the driving speed pattern by removing old driving data from the analysis target or weighting new driving data so as to greatly affect the analysis. You can also

In the above analysis, the travel environment data is not particularly mentioned, but it is also possible to select the travel data according to the travel environment data and analyze the selected travel data according to the above procedure.
Thereby, a travel speed pattern can be generated in association with the travel environment data, and the travel speed pattern can be stored in the travel speed pattern candidate storage unit 16 in association with the travel environment data.
Thereby, when setting the driving schedule of the engine 21 and the motor 24, it is possible to use the traveling speed pattern associated with the traveling environment data matching the environment at that time.

As described above, since the travel fluctuation is large in the vicinity of the predicted stop / start point such as an intersection and is an important part for analyzing the travel speed pattern, the travel pattern predicting unit 11 performs data around the predicted stop / start point. Increase the specific gravity so that the running speed pattern analysis can be made clearer.
Furthermore, the traveling pattern predicting unit 11 is traveling in a stable manner except in the vicinity of the predicted stop / start point, and it is considered that the traveling pattern predicts little influence on the analysis of the traveling speed pattern. Holding down.
Thereby, although the dispersion | variation in a driving | running | working state is large originally in the vicinity of a stop / start prediction point, it can be made to vary more than the range other than the stop / start prediction point vicinity.
In this way, by increasing the data specific gravity around the predicted stop / start point, it is possible to accurately classify traveling speed patterns having a large variation around the predicted stop / start point. In addition, the difference in fluctuation can be reduced at places where the vehicle is traveling stably (other than the vicinity of the predicted stop / start point), and the classification of the traveling speed pattern as a whole can be performed clearly and easily.

(B) A function for predicting a travel speed pattern for the travel route and setting an operation schedule for the engine 21 and the motor 24 Next, an operation schedule for the engine 21 and the motor 24 is generated using the travel speed pattern generated as described above. The function to be set will be described.
FIG. 10A is a diagram schematically showing an example of a traveling speed pattern generated for the small section 1 to the small section 4 obtained by dividing the frequent route 30 (FIG. 2). These travel speed patterns are stored in the travel speed pattern candidate storage unit 16.
These travel speed patterns are schematically shown for explaining the embodiment.

As shown in the figure, for the small section 1, there are three patterns, a traveling speed pattern 1-1 to a traveling speed pattern 1-3.
The travel speed pattern 1-1 is a pattern in which the traffic on the travel path is smooth, but deceleration and acceleration are performed in front of the predicted stop / start point. In the figure, the point where the acceleration / deceleration of the traveling speed is performed is a point before the predicted stop / start point 33a (FIG. 2).
The traveling speed pattern 1-2 is a pattern when the traffic is smooth over the entire area of the small section 1.
The travel speed pattern 1-3 is a case where the travel state of the travel path is poor and the vehicle travels while repeating deceleration and acceleration.

  For the small section 2, there is a traveling speed pattern 2-1. In the traveling speed pattern 2-1, the traveling speed is constant over the entire area of the small section. This is because, for example, the traffic signal installed at the predicted stop / start point 33a (FIG. 2) and the traffic signal installed at the predicted stop / start point 33b are linked, and after passing the predicted stop / start point 33a, the stop / start is continued. This is the case when the predicted point 33b can be passed smoothly.

For the small section 3, there are a traveling speed pattern 3-1 and a traveling speed pattern 3-2.
The traveling speed pattern 3-1 is a pattern in which deceleration and acceleration are repeated from before the predicted stop / start point 33 c (FIG. 2), and corresponds to a case where traffic around the predicted stop / start point 33 c is stagnant.
The traveling speed pattern 3-2 is a pattern for decelerating and accelerating before the predicted stop / start point 33c.
For the small section 4, there is a traveling speed pattern 4-1.
The traveling speed pattern 4-1 is a pattern in which deceleration and acceleration are repeated at a low speed as the end point is approached. For example, the traveling speed pattern 4-1 corresponds to a case where the traveling speed pattern reaches a destination.

FIG. 10B is a diagram schematically showing an example of the estimated travel speed pattern of the frequent route 30, and the travel pattern prediction unit 11 acquires the travel speed pattern of each small section from the travel speed pattern candidate storage unit 16. Are connected and generated.
In the example shown in the figure, the traveling speed pattern 1-1, the traveling speed pattern 2-1, the traveling speed pattern 3-2, and the traveling speed pattern 4-1 are selected for each of the small section 1 to the small section 4, respectively. Has been applied.

The traveling pattern prediction unit 11 generates such an estimated traveling speed pattern as follows.
First, the traveling pattern predicting unit 11 recognizes that the traveling route from now on is the frequent route 30.
For example, when the frequent route 30 is a commuting route or the like, the trains depart at approximately the same time every day, so the travel pattern prediction unit 11 determines the frequent route 30 from the current position and current time when the drive control system 10 of the hybrid vehicle is activated. It is possible to predict traveling.
Or you may comprise so that the driving | running | working path which the driver | operator will drive | work from now on may specify to the driving | running | working pattern estimation part 11 that it is the frequent route 30.

Next, the traveling pattern prediction unit 11 divides the frequent route 30 into small sections using the division point data stored in the division point storage unit 17.
Then, the travel pattern prediction unit 11 obtains one corresponding to each subsection from the travel speed patterns stored in the travel speed pattern candidate storage unit 16 and connects them along the frequent route 30.

In addition, when there are a plurality of travel speed patterns in one small section as in the small section 1, the one having the highest possibility of traveling is selected from these.
For example, it can be configured to select a traveling speed pattern associated with traveling environment data that matches the traveling environment (for example, weather, day of the week, temperature, season, etc.) on the day of driving.
Alternatively, it may be configured to select a traveling speed pattern having the largest number of section traveling data from which the traveling speed pattern is generated (that is, a traveling frequency having the highest frequency).
Or it can also comprise so that it may select using both the frequency which drive | worked and driving environment data.

FIG. 11 is a diagram schematically showing the estimated traveling speed pattern 61 and the operation schedule 62 of the engine 21 and the motor 24 generated using the estimated traveling speed pattern 61.
The estimated traveling speed pattern 61 is the same as that shown in FIG.
The estimated traveling speed pattern 61 predicts the transition of the traveling speed of the vehicle from the starting point to the destination, and the traveling pattern predicting unit 11 operates the engine 21 and the motor 24 generated based on this prediction. It is a schedule.

The driving schedule 62 is a diagram showing the transition of the SOC from the starting point to the destination planned by the traveling pattern predicting unit 11.
In the present embodiment, a schedule for handling the engine 21 and the motor 24 is set by planning the transition of the SOC.

Also in the drive control system 10 of the hybrid vehicle of this embodiment, the management range of the SOC, which is the amount of power stored in the battery 23, is set in advance, as in the case of a normal hybrid vehicle, so that the SOC is within the management range. The operation schedule is set.
The battery 23 has a voltage-current characteristic that varies depending on the SOC as in a normal battery, and the life of the battery 23 is shortened if the SOC is too large or too small. For example, when overcharged, the battery 23 may be destroyed. Therefore, for example, the management width set in advance is set so that the maximum value is about 60 [%] and the minimum value is about 40 [%], and the SOC of the battery 23 is controlled so as not to exceed the management width. The

  Therefore, for example, when driving by a motor is predicted on the previous travel route (such as uphill or low-speed driving), power is generated in an operating section where the efficiency of the engine 21 is high, or regenerative current is collected on a downhill. It is possible to plan the transition of the SOC considering the entire operation section, such as charging the battery 23 to the upper limit of the management width.

  In the driving schedule 62, the electric power is consumed in the section 71 as the vehicle starts and accelerates, and the SOC decreases, and charging is performed in the section 73 where the traveling state is stabilized. At this time, since the motor consumes power when accelerating in the section 75, the battery 23 is charged to the upper limit of the management width in the section 73.

In addition, the section 79 is a long uphill, and it is predicted that the motor 24 is operated to assist the engine 21 in this section. Therefore, the battery 23 reaches the upper limit of the management width while stably traveling in the section 77. To charge.
Furthermore, since it is predicted that the vehicle travels stably in the section 81 after the uphill in the section 79, the battery 23 is discharged to the lower limit of the management width in the section 79, and then the battery 23 is charged in the section 81.

In the subsequent stroke, the battery 23 is discharged by starting and accelerating in the section 83, and then the motor traveling at the low speed traveling in the section 87 is predicted. Therefore, the battery 23 is charged to the upper limit of the management width in the section 85. To do.
Finally, in section 87, the motor is driven so that the SOC at the time when the vehicle reaches the destination is about 50%.

  The above is a schematic explanation of an example of the generation of the operation schedule of the engine 21 and the motor 24. In this way, the transition of the traveling speed in the whole process is predicted in advance, and the operation schedule is determined based on this. By generating, the motor drive in the previous stroke is predicted and charged to the upper limit of the SOC management width in advance, or the charging opportunity in the previous stroke is predicted and discharged to the lower limit of the management width. A proactive driving schedule can be generated.

Next, an example of processing in which the travel pattern prediction unit 11 generates an operation schedule will be described with reference to FIG.
First, the traveling pattern prediction unit 11 determines whether or not the drive device of the hybrid vehicle drive control system 10 has been started (step 50).
If the drive device has not been started, the process is terminated (step 50; N), and if the hybrid vehicle drive control system 10 has been started (step 50; Y), the current position and the current time are acquired and the frequent route is obtained. It is determined whether or not traveling is predicted (step 55). Further, a frequent route that is predicted to travel from the current position and the current time is specified.
In addition, although the hybrid vehicle drive control system 10 has been described as to whether or not it is predicted that the frequent route is predicted based on the start time and the current position, and the frequent route that is predicted to travel is specified. For example, the driver may specify (specify) the frequent route, estimate the frequent route from the set destination, or perform other methods such as a route.

If frequent routes are not predicted (step 55; N), the process is terminated).
When traveling on a frequent route is predicted (step 55; Y), the traveling pattern predicting unit 11 acquires current traveling environment data from the traveling environment data acquiring unit 14 (step 60).
Next, the traveling pattern predicting unit 11 divides the frequently-occurring route on which the traveling specified above is predicted into small sections using the dividing point data stored in the dividing point storage unit 17 (step 65).

Next, the travel pattern prediction unit 11 acquires a travel speed pattern that matches the current travel environment data from the travel speed pattern candidate storage unit 16 for each divided small section (step 70).
In addition to selecting the travel speed pattern using the travel environment data, it may be selected by another method such as selecting the most frequent travel speed pattern.
Next, the traveling pattern predicting unit 11 generates the estimated traveling speed pattern of the entire traveling route by connecting the acquired traveling speed pattern for each adjacent section (step 75).
Next, the main control device 26 generates an operation schedule for the engine 21 and the motor 24 using the estimated traveling speed pattern generated in step 75 (step 80).

  As described above, in this embodiment, the engine and motor operation schedules are generated for the frequent routes. However, the present invention is not limited to this. For example, a travel speed pattern is prepared for each small section of various travel routes, and the user When the destination is set for the travel pattern prediction unit 11, the travel speed pattern of a small section passing through the travel route from the current position (departure point) to the destination is acquired and the estimated travel speed pattern is assembled. It can also be configured.

The following effects can be obtained by the present embodiment described above.
(1) Since the frequent route is divided into small sections at a point that has traveled a predetermined distance from the predicted stop / start point, the traveling state at both ends of the divided portion is relatively simplified, and changes in the vicinity of the intersections at both ends of the section Analysis can be performed without performing complicated processing for intense driving conditions.
(2) The weighted section travel data can be weighted to generate the weighted section travel data so that the transition of the change in travel speed in the variable section becomes significant.
In this way, for example, it is possible to multiply the data of a certain distance range from the predicted stop / start point such as an intersection, change the specific gravity of the data around the intersection and other than the intersection, and perform analysis with emphasis on the intersection etc. Can do.
(3) By making the weighting of data in the variable section larger than the weighting of data in the stable section, it is possible to classify the section travel data more accurately.
(4) By using the classification using the weighted section traveling data, it is possible to generate a traveling speed pattern that more accurately reflects the characteristics of each classification.
(5) An estimated traveling speed pattern can be generated using an appropriate traveling speed pattern.
(6) When applied to a hybrid vehicle, an engine and motor driving schedule can be set by using an appropriate estimated traveling speed pattern that captures the characteristics of a section with large vehicle speed fluctuations, thereby improving vehicle fuel efficiency. be able to.

It is a conceptual diagram which shows the structure of the drive control system of the hybrid vehicle in one Embodiment of this invention. In embodiment same as the above, it is a figure for demonstrating an example of the method of dividing | segmenting a frequent path | route into a subsection. In embodiment same as the above, it is a figure for demonstrating the weighting set to area travel data. It is a figure for demonstrating the area travel data and weighted area travel data in embodiment same as the above. In embodiment same as the above, it is the figure which graphed several area travel data and weighted area travel data. In embodiment same as the above, it is the figure which showed what classified heavy section travel data using the similarity of a travel data curve. In embodiment same as the above, it is the figure which showed the area travel data corresponding to the weighted area travel data after a classification | category. In embodiment same as the above, it is the figure which showed the traveling speed pattern corresponding to each classification | category. 5 is a flowchart for explaining a procedure for generating a travel speed pattern in the embodiment. In embodiment same as the above, it is a figure for demonstrating the production | generation of an estimated traveling speed pattern. In embodiment same as the above, it is a figure for demonstrating the driving schedule of the engine and motor produced | generated with respect to the estimated driving speed pattern. In embodiment same as the above, it is a flowchart for demonstrating the procedure which produces | generates a driving schedule.

Explanation of symbols

11 Travel Pattern Prediction Unit 12 Navigation DB
DESCRIPTION OF SYMBOLS 13 Travel data acquisition part 14 Travel environment data acquisition part 15 Travel data memory | storage part 16 Travel speed pattern candidate memory | storage part 17 Division | segmentation point memory | storage part 20 Drive apparatus 21 Engine 22 Generator 23 Battery 24 Motor 25 Driving force transmission apparatus 26 Main control apparatus

Claims (4)

Travel data acquisition means for acquiring and storing vehicle travel data for the travel route;
Based on the travel data, frequent route specifying means for specifying a frequent route that travels frequently,
Section dividing means for dividing the frequent route into small sections at a point traveled by a predetermined distance from a point where start and stop are predicted based on the travel data;
Classification means for classifying the travel data for each of the divided small sections for each similar travel data,
Estimated traveling speed pattern output means for generating a traveling speed pattern representing the classification for each of the small sections and outputting an estimated traveling speed pattern of a frequent route from which to travel;
A travel speed pattern estimation device comprising: The travel data acquisition means stores travel environment data in association with travel data for the travel route,
The estimated travel speed pattern output means extracts a travel speed pattern candidate that matches the current travel environment data for each of the small sections, and outputs an estimated travel speed pattern of a frequent route to be traveled from now on.
The travel speed pattern estimation apparatus according to claim 1. The section dividing means divides the frequent route into small sections at a point where the value of the predetermined distance is changed according to at least one element of the degree of traffic jam, legal speed, road width, and travel area.
The travel speed pattern estimation apparatus according to claim 1 or claim 2, wherein Drive control of a hybrid vehicle that includes an engine in which part or all of the driving force is used for driving or power generation of the vehicle and a motor that generates the driving force of the vehicle, and that travels by at least one of the driving force of the engine and the motor A device,
Power storage means that is charged by power generation and regeneration by driving the engine and supplies power to the motor;
A storage amount detection means for detecting a storage amount of the storage means;
The travel speed pattern estimation device according to claim 1, claim 2, or claim 3,
Based on the estimated traveling speed pattern of the frequent route output by the traveling speed pattern estimation device and the amount of power stored in the power storage means, the operation schedule of the engine and the motor is minimized so that the fuel consumption amount on the frequent route is minimized. An operation schedule setting means for setting
A drive control apparatus for a hybrid vehicle, comprising:

Images