DE102017215007A1 - Method for situationally controlling a vehicle fleet system - Google Patents

Abstract

The present invention includes a method for situationally controlling a vehicle fleet system. The method includes receiving at a back-end server location data of a plurality of vehicles associated with the vehicle fleet system. The method comprises determining, via an evaluation unit of the back-end server, a situational vehicle requirement for a multiplicity of partial areas of a predefined area to which the vehicle fleet system is assigned, wherein at least a part of the evaluation unit is designed as a neural network, such that the determination of the situational Vehicle demand dynamically based on a variety of influence parameters for each sub-area. The method includes initiating at least one measure for reorganizing unused vehicles such that the number of vehicles in each subarea is adjusted according to the determined demand.

Description

The present invention relates to a method for situationally controlling a vehicle fleet system.

Vehicle fleet systems, for example car sharing systems, are known. In particular, it is known that car sharing operators assign an area in a city to a business area. This business area is assigned a predetermined number of vehicles. The number of vehicles is usually determined by a statically calculated need for car sharing vehicles in each area. A disadvantage of this procedure is that the static demand determination is inflexible with regard to situation-dependent fluctuations in demand, for example at rush hours and / or on main traffic routes.

The object of the invention is to provide a solution which allows a situational control of a vehicle fleet system, so that dynamic adjustments to situation-dependent demand fluctuations are made possible.

This object is achieved by the features of the independent claim. Preferred embodiments are subject of the dependent claims.

• Empfangen, an einem Backend-Server, von Positionsdaten einer Vielzahl von Fahrzeugen, die dem Fahrzeugflottensystem zugeordnet sind;
• Ermitteln, über eine Auswerteeinheit des Backend-Servers, eines situativen Fahrzeugbedarfs für eine Vielzahl von Teilbereichen eines vordefinierten Bereichs, dem das Fahrzeugflottensystem zugeordnet ist,
• wobei wenigstens ein Teil der Auswerteeinheit als neuronales Netz ausgebildet ist, so dass das Ermitteln des situativen Fahrzeugbedarfs dynamisch anhand einer Vielzahl an Einflussparametern für jeden Teilbereich erfolgt; und
• Einleiten zumindest einer Maßnahme zum Umorganisieren nicht benutzter Fahrzeuge derart, dass die Anzahl der Fahrzeuge in jedem Teilbereich entsprechend des ermittelten Bedarfs angepasst wird.

The above object is achieved by a method for situationally controlling a vehicle fleet system, comprising:

• Receiving, at a backend server, location data of a plurality of vehicles associated with the vehicle fleet system;
• Determining, via an evaluation unit of the backend server, a situational vehicle requirement for a multiplicity of subareas of a predefined area to which the vehicle fleet system is assigned,
• wherein at least a part of the evaluation unit is designed as a neural network, so that the determining of the situational vehicle demand takes place dynamically on the basis of a multiplicity of influencing parameters for each subarea; and
• Initiate at least one measure for reorganizing unused vehicles such that the number of vehicles in each subarea is adjusted according to the determined demand.

The term vehicle includes in particular passenger cars (cars), trucks (trucks), buses, campers, motorcycles, etc.

The backend server is a central data pool and may include a computing device, as well as a memory device, e.g. a database, in which data can be centrally or centrally controlled and stored, managed and processed outside the vehicle. It may be necessary for each vehicle to first register at the backend server (or other suitable computing device providing a corresponding service). The one-time registration may include the deposit of a suitable vehicle identification (ID). In this way, data sent to the backend server by the respective vehicle can be uniquely assigned to it.

In addition, it can be stored at the back-end server for each vehicle, which vehicle fleet is assigned to this and which area or which geographical area is assigned to the vehicle fleet system. The backend server, the vehicles and the one or more vehicle fleets to which the vehicles are associated are part of the vehicle fleet system.

The back-end server comprises an evaluation unit, wherein at least a part of the evaluation unit is designed as an artificial neural network. Neural networks form part of machine learning, thus generating knowledge from experience. To ensure optimum vehicle fleet system vehicles deployment, the neural network performs dynamic demand determination. For this purpose, a demand determination model is used in the neural network, which is displayed on the basis of certain influencing parameters and thus characterizes the vehicle requirement for each subarea of the area or geographical area assigned to the vehicle fleet. For this, the influencing parameters are mapped into the neural network. In addition, the demand determination model can be expanded by also mapping the respective current actual state of the geographical area or area assigned to the vehicle fleet, for example changes in the road guidance, in public transport, etc., into the neural network as well. The demand determination model may be implemented in the neural network in a known manner using one or more suitable learning algorithms.

The back-end server is configured to receive current position data from the vehicles associated with the vehicle fleet system. The current position or current position data of the vehicle can be determined by means of a navigation satellite system. The Navigation Satellite System can be any current or future Global Navigation Satellite System (GNSS) for position determination and navigation by receiving the signals from navigation satellites and / or pseudolites act. For example, these may be Global Positioning System (GPS), GLOBAL NAvigation Satellite System (GLONASS), Galileo, positioning system, and / or BeiDou Navigation Satellite System. For example, the vehicle may comprise a module which is suitable for detecting a position of the vehicle in the respective system. In the example of GPS, the vehicle may include a position detection unit comprising a GPS module configured to determine the current GPS position of the vehicle. The position data is finally sent to the backend server.

Each vehicle may include a communication module. The communication module may be a communication module arranged in the vehicle, which is able to establish a communication connection with other communication participants, for example with the back-end server. The communication module may comprise a subscriber identity module or a SIM card, which serves to establish a communication connection via a mobile radio system. The subscriber identity module uniquely identifies the communication module in the mobile radio network. The communications link may be a data link (e.g., packet switch) and / or a wired communications link (e.g., circuit switch). Any communication between the vehicle and other communication participants may be via the communication module.

The vehicles can capture the position data on predefined events and transmit them to the backend server. This can be done at regular intervals, e.g. every 30 seconds, every minute, every 2, 5, 10 or 20 minutes. In addition, or alternatively, any vehicle may be arranged to transmit the position data to predefined events, e.g. at every journey, at every stop, at every parking, when leaving the predefined area assigned to the vehicle fleet system, etc.

The vehicles belong to the vehicle fleet system to which at least one geographical area is assigned. For example, the associated geographic area may be a business area of a vehicle fleet of a car sharing provider in a city where the vehicles can be moved and parked at no additional cost. The geographical area is divided into subareas.

The evaluation unit can now use the demand determination model implemented in the neural network to determine the situational vehicle requirement for each subarea of the geographical area assigned to the vehicle fleet system or the vehicle fleet. Determining the situational vehicle requirements for the subregions can take place at regular time intervals and / or at predefined events. For example, the determination can be made every 30, 60, 90, minutes, etc. In addition or alternatively, the determination can take place if no free vehicle is available in at least one partial area, etc.

The back-end server is configured to carry out or implement at least one measure for reorganizing vehicles that are not used at the time such that the number of vehicles in each sub-area is adjusted in accordance with the determined vehicle requirement in the respective sub-areas.

A vehicle may be considered unused if it is currently not being moved or reserved by any user of the vehicle fleet system. The adaptation to the ascertained vehicle requirement in the respective subareas means that unused vehicles are distributed to the subareas in such a way that the subdivision best corresponds to the determined vehicle requirement. The breakdown can be implemented taking into account economic factors, such as travel costs and non-autonomous vehicles personnel expenses.

Preferably, at least a part of the vehicles has a fully autonomous driving mode, wherein the measure comprises controlling the vehicles comprising a fully autonomous driving mode such that they divide onto the subareas in accordance with the determined demand.

Vehicles comprising a fully autonomous driving mode are known. Such vehicles can move autonomously by means of a driver assistance device or an autopilot in traffic. Thus, the driving of the vehicle can be performed independently of a user or driver of the vehicle, so that when activated fully autonomous driving mode no driver needs to be in the vehicle. While the fully autonomous driving mode is active, the driver assistance device continuously processes current environmental data. For this purpose, a plurality of different sensors mounted in or on the vehicle to capture environmental data. The sensors may be, for example, ultrasonic sensors, lidar sensors, radar sensors, cameras, etc. The sensed sensor data is then processed in the driver assistance device (or other suitable computing device) to determine an environmental model of the vehicle. On Based on the environment model, the fully autonomous driving mode is executed.

The back-end server can be set up to control vehicles comprising a fully autonomous driving mode in such a way that they optimally split up into the subregions in accordance with the determined vehicle requirement. For this purpose, the back-end server can transmit a corresponding message to the communication module of the vehicle. The message may include a request to position at an optimal geographic location in a subarea. The optimal geographic location may be communicated with the message, such as GPS location data describing the geographic location. A suitable computing unit of the vehicle can process the message, and the driver assistance module can control the vehicle such that it autonomously moves to the optimal geographical position in the corresponding partial area.

• - eine Anzahl aktiver Nutzer von Fahrzeugen des Fahrzeugflottensystems;
• - eine Anzahl zum Zeitpunkt nicht betriebener Fahrzeuge;
• - Informationen zu Buchungen und Buchungsversuchen der Nutzer von Fahrzeugen des Fahrzeugflottensystems;
• - aktuelle Wetterdaten, wie beispielsweise Temperatur und Niederschlagsmenge;
• - aufgezeichnete Bewegungsdaten von Fahrzeugen des Fahrzeugflottensystems;
• - Informationen zum vergangenen, aktuellen und voraussichtlichen Verkehrsaufkommen;
• - Informationen zur vergangenen, aktuellen und voraussichtlichen Parksituation in den jeweiligen Teilbereichen;
• - Daten zu vergangenen, aktuellen und kommenden Veranstaltungen, wie beispielsweise Sportveranstaltungen, Konzerte oder Messeveranstaltungen;
• - Standpunkte von Haltestellen, Verlauf, Abfahrtszeiten und Typ von öffentlichen Verkehrsmitteln, wie beispielsweise Bussen, U-Bahnen, Straßenbahnen und Zügen;
• - Standpunkte von besonderen Zielorten, wie beispielsweise Sehenswürdigkeiten, Restaurants oder Einkaufszentren;
• - Daten zu voraussichtlichen Zielorten von Fahrzeugen des Fahrzeugflottensystems;
• - sonstige Parameter, die den Fahrzeugbedarf beeinflussen können.

Preferably, the influencing parameters include:

• a number of active users of vehicles of the vehicle fleet system;
• - a number of non-powered vehicles at the time;
• - information on bookings and booking attempts of users of vehicles of the vehicle fleet system;
• - current weather data, such as temperature and rainfall;
• - recorded movement data of vehicles of the vehicle fleet system;
• - information on past, present and anticipated traffic volumes;
• - Information on the past, current and prospective parking situation in the respective subareas;
• - data on past, current and upcoming events, such as sporting events, concerts or trade fairs;
• - Location of stops, history, departure times and type of public transport, such as buses, subways, trams and trains;
• - Points of view of specific destinations, such as attractions, restaurants or shopping centers;
• - data on expected destinations of vehicles of the vehicle fleet system;
• - other parameters that may influence vehicle requirements.

The measure preferably comprises a redistribution of vehicles between different vehicle fleets.

For example, a car sharing provider may offer one vehicle fleet each in a specific geographic area in multiple cities. The vehicle demand determination may show that the vehicle requirement is at least temporarily shifts so that it makes sense to redistribute vehicles of a vehicle fleet of a first city in the vehicle fleet of a second city.

Preferably, the neural network is a recurrent neural network.

A recurrent neural network is an artificial neural network in which connections between neurons are present not only from one layer to a next layer, but also connections between neurons of the same layer and / or a previous layer. Advantageously, temporally coded information in the data can be identified by the recurrent or feedback connection of the neurons. The recurrent neural network may be, for example, the implementation of the Elman network, the Jordan network or the Hopfield network.

• 1 zeigt ein schematisches System, das geeignet ist, ein Verfahren zum situativen Steuern eines Fahrzeugflottensystems durchzuführen;
• 2 zeigt eine beispielhafte Einteilung eines geografischen Bereichs in zwei Teilbereiche;
• 3 zeigt ein Flussdiagramm, das ein Verfahren zum situativen Steuern eines Fahrzeugflottensystems veranschaulicht.

These and other objects, features and advantages of the present invention will become more apparent from a study of the following detailed description of preferred embodiments and the accompanying drawings. It will be appreciated that although embodiments are described separately, individual features thereof may be combined to form additional embodiments.

• 1 shows a schematic system that is adapted to perform a method for situationally controlling a vehicle fleet system;
• 2 shows an exemplary division of a geographical area into two partial areas;
• 3 FIG. 12 is a flowchart illustrating a method for situationally controlling a vehicle fleet system.

1 shows schematic vehicle fleet system 100 on which a procedure 300 for situational control of a vehicle fleet system 100 can be carried out. The procedure 300 becomes with reference to below 2 and 3 exemplified in more detail.

The system 100 includes at least one backend server 120 , The backend server 120 is a central data pool and can be a computing device 122 , as well as a memory device 125 , eg a database, in which data can be centrally or centrally controlled and stored, managed and processed outside the vehicle. It may be necessary for every vehicle 110 first at the backend server 120 (or any other suitable computing device that provides a corresponding service) a one-time registration takes place. The one-time registration may include the deposit of a suitable vehicle identification (ID). So can data following from the respective vehicle 110 to the backend server 120 be assigned to this unique.

In addition, on the backend server 120 for every vehicle 110 be deposited (eg during the registration), which vehicle fleet this is assigned and which area or which geographical area of the respective vehicle fleet is assigned. For example, a car sharing operator can provide one vehicle fleet per city. Each vehicle fleet can be assigned to vehicles as well as a geographical area or area. The backend server 120 , the vehicles 110 as well as the one or more vehicle fleets to which the vehicles 110 are assigned, are part of the vehicle fleet system 100 ,

The backend server 120 includes an evaluation unit 124 , wherein at least a part of the evaluation unit 124 is designed as an artificial neural network. In this example, the evaluation unit 124 can as part of the computing device 122 or realized as a separate arithmetic unit.

Neural networks are a part of machine learning and generate knowledge from experience. To ensure optimal provision of the vehicles 110 of the vehicle fleet system 100 A dynamic demand determination is performed using the neural network. For this purpose, a demand determination model is used in the neural network, which is displayed on the basis of certain influencing parameters and thus the vehicle requirements for each subarea 210 . 220 (see. 2 ) of the at least one vehicle fleet assigned area or geographical area 200 characterized. For this, the influencing parameters are mapped into the neural network. In addition, the demand determination model can be expanded by the respective current actual state of the geographical area or area assigned to the vehicle fleet 200 For example, changes in the road, in public transport, etc., is also mapped into the neural network. The demand determination model may be implemented in the neural network in a known manner using one or more suitable learning algorithms.

The neural network can be a recurrent neural network. A recurrent neural network is an artificial neural network in which connections between neurons are present not only from one layer to a next layer, but also connections between neurons of the same layer and / or a previous layer. Advantageously, temporally coded information in the data can be identified by the recurrent or feedback connection of the neurons. The recurrent neural network may be, for example, the implementation of the Elman network, the Jordan network or the Hopfield network.

The backend server 120 is set up position data from a variety of vehicles 110 that the vehicle fleet system 100 are assigned to receive 210 , In particular, the backend server 120 set up, current position data from the vehicles 110 to recieve. The current position or the current position data of a vehicle 110 can be determined using a navigation satellite system. The navigation satellite system can be any common and future global navigation satellite system (GNSS) for position determination and navigation by receiving the signals from navigation satellites and / or pseudolites. For example, these may be Global Positioning System (GPS), GLOBAL NAvigation Satellite System (GLONASS), Galileo, positioning system, and / or BeiDou Navigation Satellite System. For example, any vehicle 110 comprise a module which is suitable for detecting a position of the vehicle in the respective system. In the example of GPS can be a vehicle 110 a position determination unit 114 comprising a GPS module (not shown), which is configured, the current GPS position or current GPS position data of the vehicle 110 to investigate. The position data are from each (possibly previously registered) vehicle 110 to the backend server 120 Posted.

Every vehicle 110 can be a communication module 112 include. The communication module 112 can one in the vehicle 110 arranged communication module 112 which is capable of communicating with other communication parties, such as the backend server 120 to build up. The communication module 112 can be a subscriber identity module or a Subscriber Identity Module or a SIM card (not shown), which (s) serves to establish a communication connection via a mobile radio system. The subscriber identity module identifies the communication module 112 clearly in the mobile network. The communication connection may be a data connection (eg packet switching) and / or a wired communication connection (eg circuit switching). Any communication between the vehicle 110 and other communication participants can via the communication module 112 respectively.

The vehicles 110 can capture the position data about predefined events and to the backend server 120 to transfer. This can be done at regular intervals, eg every 30 seconds, every minute, every 2, 5, 10 or 20 minutes. In addition, or alternatively, any vehicle 110 be set up to transmit the position data on predefined events, for example, at each departure, at every stop, at each parking, when leaving the predefined, the vehicle fleet system 110 assigned area, etc.

The vehicle fleet system 100 includes at least one vehicle fleet. Every vehicle 100 is assigned to a vehicle fleet. Each vehicle fleet is a geographical area 200 assigned as below with reference to 2 explained in more detail. The associated geographic area 200 is in subareas 210 . 220 divided up.

The backend server 120 is set up via the evaluation unit 124 a situational vehicle requirement for the subareas 210 . 220 the predefined geographic area 200 which the vehicle fleet system 100 (or in the case of several vehicle fleets, a vehicle fleet of the vehicle fleet system 100 ) is determined to determine. Because at least part of the evaluation unit 124 is designed as a neural network, the determination is done 320 the situational vehicle demand dynamically based on a variety of influence parameters for each sub-area 210 . 220 , The evaluation unit 124 can now use the needs assessment model implemented in the neural network to determine the situational vehicle requirements for each subarea 210 of the vehicle fleet system 100 associated geographic area 200 determine. Determining 320 the situational vehicle requirements for the subareas 210 . 220 may be at regular intervals and / or predefined events. For example, the determination can be made every 30, 60, 90, minutes, etc. In addition or alternatively, the determination can take place if no free vehicle is available in at least one partial area, etc.

• - eine Anzahl aktiver Nutzer von Fahrzeugen des Fahrzeugflottensystems;
• - eine Anzahl zum Zeitpunkt nicht betriebener Fahrzeuge;
• - Informationen zu Buchungen und Buchungsversuchen der Nutzer von Fahrzeugen des Fahrzeugflottensystems;
• - aktuelle Wetterdaten, wie beispielsweise Temperatur und Niederschlagsmenge;
• - aufgezeichnete Bewegungsdaten von Fahrzeugen des Fahrzeugflottensystems;
• - Informationen zum vergangenen, aktuellen und voraussichtlichen Verkehrsaufkommen;
• - Informationen zur vergangenen, aktuellen und voraussichtlichen Parksituation in den jeweiligen Teilbereichen;
• - Daten zu vergangenen, aktuellen und kommenden Veranstaltungen, wie beispielsweise Sportveranstaltungen, Konzerte oder Messeveranstaltungen;
• - Standpunkte von Haltestellen, Verlauf, Abfahrtszeiten und Typ von öffentlichen Verkehrsmitteln, wie beispielsweise Bussen, U-Bahnen, Straßenbahnen und Zügen;
• - Standpunkte von besonderen Zielorten, wie beispielsweise Sehenswürdigkeiten, Restaurants oder Einkaufszentren;
• - Daten zu voraussichtlichen Zielorten von Fahrzeugen des Fahrzeugflottensystems;
• - sonstige Parameter, die den Fahrzeugbedarf beeinflussen können.

The influencing parameters may include one or more of the following, and may be dynamically augmented or adapted to new findings and / or events that may affect vehicle needs:

• a number of active users of vehicles of the vehicle fleet system;
• - a number of non-powered vehicles at the time;
• - information on bookings and booking attempts of users of vehicles of the vehicle fleet system;
• - current weather data, such as temperature and rainfall;
• - recorded movement data of vehicles of the vehicle fleet system;
• - information on past, present and anticipated traffic volumes;
• - Information on the past, current and prospective parking situation in the respective subareas;
• - data on past, current and upcoming events, such as sporting events, concerts or trade fairs;
• - Location of stops, history, departure times and type of public transport, such as buses, subways, trams and trains;
• - Points of view of specific destinations, such as attractions, restaurants or shopping centers;
• - data on expected destinations of vehicles of the vehicle fleet system;
• - other parameters that may influence vehicle requirements.

The backend server 120 is also set up, at least one measure to reorganize unused vehicles 110 include. The reorganization of unused vehicles 110 is done such that the number of vehicles 110 in every section 210 . 220 adapted to the determined vehicle needs.

A vehicle 110 can not be considered as being used if it is currently not owned by a user of the vehicle fleet system 100 moved, is still reserved. The adaptation to the determined vehicle requirement in the respective subareas 210 . 220 includes that unused vehicles are divided into the sub-areas such that the distribution of the identified vehicle needs most optimally. The division may be under consideration economic factors, such as travel costs and non-autonomous vehicles 110 Personnel expenses, to be implemented.

At least part of the vehicles 110 can have a fully autonomous driving mode. vehicles 110 that include a fully autonomous driving mode are known. Such vehicles 110 can autonomously by means of a driver assistance device or an autopilot 116 moving in traffic. Thus, driving the vehicle 110 independent of a user or driver of the vehicle 110 respectively. While the fully autonomous driving mode is active, the driver assistance device processes 116 continuously updated environmental data. This is a variety of different sensors 118 in or on the vehicle 110 attached to capture environmental data. With the sensors 118 These may be, for example, ultrasonic sensors, lidar sensors, radar sensors, cameras, etc. The detected sensor data are then in the driver assistance device 118 processed to an environment model of the vehicle 110 to investigate. Based on the environment model, the fully autonomous driving mode is executed.

In this example, the measure may be controlling the vehicles 110 comprising a fully autonomous driving mode such that they correspond to the determined needs of the subregions 210 . 220 split up. The backend server 120 can be set up vehicles 110 comprising a fully autonomous driving mode to control such that they optimally according to the determined vehicle demand on the sub-areas 210 . 220 split. This can be done by the backend server 120 an appropriate message to the communication module 112 of the vehicle 110 to transfer. The message may prompt for positioning at an optimal geographical position in a subarea 210 include. The optimal geographic location may be communicated with the message, such as GPS location data describing the geographic location. A computing unit (not shown) of the vehicle 110 can process the message and the driver assistant device 116 can the vehicle 110 such that it autonomously adjusts to the optimal geographic position in the corresponding subarea 210 . 220 emotional. Advantageously, so parked or unused vehicles 110 comprehensively moving a fully autonomous driving mode away from the dangerous situation, whereby the danger potential around the dangerous situation is further reduced.

In addition, or alternatively, the action may involve redistributing vehicles 110 between different vehicle fleets that the vehicle fleet system 100 are associated.

For example, a car sharing provider may offer multiple vehicle fleets in multiple cities, with the vehicle fleets being a single vehicle fleet system 100 assigned. In this example, each vehicle fleet becomes a geographic area 200 , each in subareas 210 . 220 is subdivided. The vehicle demand determination can show that the vehicle requirement is at least temporarily shifts so that it makes sense vehicles 110 to redistribute or allocate a vehicle fleet of a first city into the vehicle fleet of a second city. The redistribution can be done by vehicles 110 , which have a fully autonomous driving mode, in the aforementioned way automatically by a message from the backend server 120 respectively.

The respective units or modules 112 . 114 . 116 as well as the sensors 118 of the vehicle can be connected to each other at least communicatively, for example via a bus system for data transmission. Examples of a bus system or bus connection are FlexRay (FR), Controller Area Network (CAN), Local Interconnect Network (LIN) and Ethernet.

2 shows an exemplary geographical area 200 , the vehicle fleet of the vehicle fleet system 100 as related to 1 and 3 explained in more detail, is assigned. In particular, a schematic plan of a city area becomes 200 shown in two subsections 210 . 220 is divided. The subdivision is based on the double line, which is the geographical area 200 pervades, evident.

The vehicle fleet system 100 includes at least one vehicle fleet. Every vehicle 100 is assigned to a vehicle fleet. Each vehicle fleet is a geographical area 200 assigned. The associated geographic area 200 may for example be a business area of a car sharing provider in a city where the vehicles can be moved and parked at no additional cost. The division of the geographical area 200 in two parts 210 . 220 is just an example. Depending on the size of the geographical area 200 is to choose a meaningful subdivision into subareas.

• - eine Anzahl aktiver Nutzer von Fahrzeugen 110 des Fahrzeugflottensystems;
• - eine Anzahl zum Zeitpunkt nicht betriebener Fahrzeuge 110;
• - Informationen zu Buchungen und Buchungsversuchen der Nutzer von Fahrzeugen 110 des Fahrzeugflottensystems;
• - aktuelle Wetterdaten, wie beispielsweise Temperatur und Niederschlagsmenge;
• - aufgezeichnete Bewegungsdaten von Fahrzeugen 110 des Fahrzeugflottensystems;
• - Informationen zum vergangenen, aktuellen und voraussichtlichen Verkehrsaufkommen;
• - Informationen zur vergangenen, aktuellen und voraussichtlichen Parksituation in den jeweiligen Teilbereichen 210, 220;
• - Daten zu vergangenen, aktuellen und kommenden Veranstaltungen, wie beispielsweise Sportveranstaltungen, Konzerte oder Messeveranstaltungen;
• - Standpunkte von Haltestellen, Verlauf, Abfahrtszeiten und Typ von öffentlichen Verkehrsmitteln, wie beispielsweise Bussen, U-Bahnen, Straßenbahnen und Zügen;
• - Standpunkte von besonderen Zielorten, wie beispielsweise Sehenswürdigkeiten, Restaurants oder Einkaufszentren;
• - Daten zu voraussichtlichen Zielorten von Fahrzeugen 110 des Fahrzeugflottensystems;
• - sonstige Parameter, die den Fahrzeugbedarf beeinflussen können.

3 shows a flowchart illustrating a method 300 for situational control of a vehicle fleet system 100 as with respect to the 1 and 2 described. The method includes receiving 310 , on a backend server 120 , positional data of a variety of vehicles 110 that the vehicle fleet system 100 assigned. At least part of the vehicles 110 can have a fully autonomous driving mode. The method also includes determining 320 , via an evaluation unit 124 the backend server 120 , a situational vehicle needs for a variety of sub-areas 210 . 220 a predefined area 200 to which the vehicle fleet system is assigned, wherein at least a part of the evaluation unit 124 is designed as a neural network, so that the determining 320 the situational vehicle demand dynamically based on a variety of influence parameters for each sub-area 210 . 220 he follows. The neural network can be designed as a recurrent neural network. The influencing parameters may include:

• - a number of active users of vehicles 110 the vehicle fleet system;
• - a number at the time of non-powered vehicles 110 ;
• - Information about bookings and booking attempts of users of vehicles 110 the vehicle fleet system;
• - current weather data, such as temperature and rainfall;
• - recorded movement data of vehicles 110 the vehicle fleet system;
• - information on past, present and anticipated traffic volumes;
• - Information on the past, current and prospective parking situation in the respective subareas 210 . 220 ;
• - data on past, current and upcoming events, such as sporting events, concerts or trade fairs;
• - Location of stops, history, departure times and type of public transport, such as buses, subways, trams and trains;
• - Points of view of specific destinations, such as attractions, restaurants or shopping centers;
• - Data on expected destinations of vehicles 110 the vehicle fleet system;
• - other parameters that may influence vehicle requirements.

The procedure 300 also includes an initiation 330 at least one measure to reorganize unused vehicles 110 such that the number of vehicles 110 in every section 210 . 220 adjusted according to the identified needs.

In the case where at least part of the vehicles 110 a fully autonomous driving mode, the measure may include controlling unused vehicles 110 comprising a fully autonomous driving mode such that they correspond to the determined needs of the subregions 210 . 220 split up.

In addition, or alternatively, the action may involve redistributing vehicles 110 between different vehicle fleets.

Claims (5)

A method (300) for situationally controlling a vehicle fleet system, comprising: Receiving (310), at a backend server (120), positional data of a plurality of vehicles (110) associated with the vehicle fleet system; Determining (320), via an evaluation unit (124) of the back-end server (120), a situational vehicle requirement for a plurality of partial areas (210, 220) of a predefined area (200) to which the vehicle fleet system is assigned, wherein at least a part of the evaluation unit (124) is designed as a neural network, so that the determining (320) of the situational vehicle demand takes place dynamically on the basis of a plurality of influencing parameters for each partial area (210, 220); and Initiating (330) at least one measure to reorganize unused vehicles such that the number of vehicles (110) in each subarea (210, 220) is adjusted according to the determined demand. Method (300) according to Claim 1 wherein at least a portion of the vehicles (110) have a fully autonomous driving mode, and wherein the measure comprises controlling unused vehicles (110) including a fully autonomous driving mode such that they divide into the subregions (210, 220) according to the determined demand. Method (300) according to Claim 1 or 2 wherein the influencing parameters comprise: a number of active users of vehicles (110) of the vehicle fleet system; a number of non-powered vehicles (110) at the time; - information on bookings and booking attempts of the users of vehicles (110) of the vehicle fleet system; - current weather data, such as temperature and rainfall; - recorded movement data of vehicles (110) of the vehicle fleet system; - information on past, present and anticipated traffic volumes; - Information on the past, current and prospective parking situation in the respective subareas (210, 220); - Data on past, current and upcoming events, such as sporting events, concerts or trade fairs; - Location of stops, history, departure times and type of public transport, such as buses, subways, trams and trains; - Points of view of specific destinations, such as attractions, restaurants or shopping centers; - data on expected destinations of vehicles (110) of the vehicle fleet system; and / or - other parameters that may affect vehicle needs. Method (300) according to Claim 2 or 3 wherein the measure comprises redistributing vehicles (110) between different vehicle fleets. The method (300) according to one of the preceding claims, wherein the neural network is a recurrent neural network.

Images