TW202409989A - Evaluation method of locations, analysis method of driving behavior, and driver management system - Google Patents

Abstract

An evaluation method of locations, an analysis method of driving behavior, and a driver management system are provided. In the method, sensing data is obtained. The parking state is determined according to the sensing data. A parking location category corresponding to the sensing data under the parking state is determined. A location suggestion model is trained according to the parking location category and the sensing data. The location suggestion model is used for a suggestion of a parking location. The location suggestion model is trained through a machine learning algorithm. An energy-saving score of the sensing data is determined according to one or more energy-saving factors. The driving behavior report is generated according to the energy-saving scores. The energy-saving factor is a factor that affects the energy consumption of the car. The driving behavior report describes good and bad of the energy-saving score. Accordingly, the training and work efficiency could be improved.

Description

Location evaluation method, driving behavior analysis method and driving management system

The present invention relates to a data analysis technology, and in particular, to a location evaluation method, a driving behavior analysis method and a driving management system.

Working conditions in the fleet industry have resulted in problems such as an aging workforce and high driver turnover. In recent years, the pandemic has dramatically increased demand for the global logistics industry, requiring the hiring of more new drivers to handle a large number of delivery operations. Therefore, the importance of improving the education and training of novice drivers is not only related to driver safety, but also affects fleet operations (for example, costs due to fines or vehicle damage).

The driver shortage problem is becoming increasingly serious. Some countries have even opened up to foreign drivers to fill the gap. When drivers arrive in a new country, they may be unfamiliar with the traffic conditions, resulting in low delivery efficiency. If the delivery range is in an urban area, it is even more difficult for novice drivers to find a place to park and deliver the goods.

In view of this, the embodiments of the present invention provide a location evaluation method, a driving behavior analysis method and a driving management system, which can automatically recommend parking locations and determine the quality of driving behavior.

The location evaluation method of the embodiment of the present invention includes (but is not limited to) the following steps: obtaining sensing data; determining a parking state based on the sensing data; obtaining a parking location category corresponding to the sensing data in the parking state; and training a location recommendation model based on the parking location category and the sensing data. The location recommendation model is used to recommend parking locations.

The driving behavior analysis method of the embodiment of the present invention includes (but is not limited to) the following steps: obtaining sensing data; determining an energy-saving score of the sensing data based on one or more energy-saving factors; and generating a driving behavior report based on the energy-saving score. The energy-saving factor is a factor that affects the energy consumption of the vehicle. The driving behavior report describes the pros and cons of the energy-saving score.

The driving management system of the embodiment of the present invention includes a server. The server communication is connected to the vehicle-mounted device. The server obtains the parking location category corresponding to the sensing data of the vehicle-mounted device in the parking state, and trains the location recommendation model based on the parking location category and the sensing data. The location suggestion model is used to recommend parking locations.

The driving management system of the embodiment of the present invention includes a server. The server communication is connected to the vehicle-mounted device. The server determines the energy saving score of the sensing data of the vehicle-mounted device based on one or more energy saving factors, and generates a driving behavior report based on the energy saving score. The energy saving factor is a factor that affects the energy consumption of a vehicle equipped with an in-vehicle device, and the driving behavior report explains the advantages and disadvantages of the energy saving score.

Based on the above, according to the location evaluation method, driving behavior analysis method and driving management system of the embodiments of the present invention, sensing data can be collected to recommend parking locations and/or evaluate driving behavior. In this way, the efficiency of task execution and driving safety can be improved, and management of fleet operators can be facilitated.

In order to make the above features and advantages of the present invention more clearly understood, embodiments are specifically cited below and described in detail with reference to the accompanying drawings.

Figure 1 is a schematic diagram of a driving management system 1 according to an embodiment of the present invention. Referring to FIG. 1 , the driving management system 1 includes one or more vehicle-mounted devices 10 , a network access device 20 , a server 30 and one or more remote devices 40 .

The vehicle-mounted device 10 is an electronic device that is loaded, mounted, placed or built into a vehicle. The network access device 20 (as shown in FIG. 1 ) may be a base station, a router, a relay station, a core network device or a combination thereof. In one embodiment, the vehicle-mounted device 10 is connected to the Internet, a local area network or a private network via the network access device 20. In another embodiment, the vehicle-mounted device 10 may communicate directly with other devices. The server 30 may be a smartphone, a tablet computer, a desktop computer, a laptop computer or a cloud platform device. The remote device 40 (as shown in FIG. 1 ) may be a smartphone, a tablet computer, a desktop computer, a laptop computer, a wearable device or an intelligent assistant device.

FIG. 2 is a component block diagram of the vehicle-mounted device 10 according to an embodiment of the present invention. Referring to FIG. 2 , the vehicle-mounted device 10 includes (but is not limited to) one or more sensors 11 , a positioning device 12 , a communication transceiver 13 , a display 14 , a memory 15 and a processor 16 .

The sensor 11 may be an image sensor, an accelerometer, a gyroscope, an electronic compass, an inertial sensor, an on-board diagnostics system (OBD), and/or a thermometer. In one embodiment, the sensor 11 is used to obtain sensing data. The sensing data may be images, acceleration, speed, direction, angular velocity, vehicle speed, engine speed, and/or braking behavior.

The positioning device 12 may support reception of the Global Positioning System (GPS), GLONASS, GALILEO, BeiDou Navigation Satellite System or other satellite positioning systems. device. For example, obtain a 10MHz and 1PPS (One Pulse Per Second) precise time source through an ANT antenna. In one embodiment, the positioning device 12 is used to receive positioning signals and generate location information and/or time information accordingly. The location information can be latitude and longitude, coordinates in other geographical coordinate systems, or relative position. Time information may include time zone, time and/or date. For example, National Marine Electronics Association (NMEA) information. In some embodiments, one or more pieces of location information and their corresponding time information can be recorded as a positioning record.

The communication transceiver 13 may be a transceiver that supports wireless communication technology such as Wi-Fi, mobile network, or Bluetooth, or a transceiver that supports wired communication technology such as Ethernet, optical fiber network, or USB. In one embodiment, the communication transceiver 13 is used to transmit or receive data with external electronic devices (for example, the network access device 20, the server 30 or the remote device 40).

The display 14 (optional) may be a liquid crystal display (Liquid-Crystal Display, LCD), a (Light-Emitting Diode, LED) display, an organic light-emitting diode (Organic Light-Emitting Diode, OLED), or a quantum dot display (Quantum dot display) or other type of display. In one embodiment, the display 14 is used to display images or user interfaces.

The memory 15 can be any type of fixed or removable random access memory (Random Access Memory, RAM), read-only memory (Read-Only Memory, ROM), flash memory (Flash Memory) or similar components or a combination of the above elements. In one embodiment, the memory 15 stores program codes, device configurations, codebooks, temporary or permanent data (for example, sensing data, positioning information, energy saving scores or driving behavior reports).

The processor 16 may be a central processing unit (CPU), a graphics processing unit (GPU), or other programmable general-purpose or special-purpose microprocessor, digital signal processor (DSP), programmable controller, field programmable gate array (FPGA), application-specific integrated circuit (ASIC), neural network accelerator or other similar components or a combination of the above components. In one embodiment, the processor 16 is used to execute all or part of the operations of the vehicle-mounted device 10. In some embodiments, the functions of the processor 16 can be implemented through software or chips.

In some embodiments, one or more components in the vehicle-mounted device 10 can be separated into other independent devices (eg, Internet of Things (IoT) devices, driving recorders, or smartphones), and these devices can form a vehicle-mounted system. Each device in the vehicle system can send or receive data to each other through the communication transceiver 13 .

Please refer to FIG. 3 , which is a component diagram of the server 30 according to an embodiment of the present invention. The server 30 includes (but is not limited to) a communication transceiver 33 , a memory 35 and a processor 36 .

The implementation and functions of the communication transceiver 33, the memory 35 and the processor 36 can be respectively referred to the communication transceiver 13, the memory 15 and the processor 16 in FIG. 2, and will not be repeated here. In one embodiment, the processor 36 is used to execute all or part of the operations of the server 30.

In the following, the method described in the embodiment of the present invention will be explained in conjunction with each device and/or component in the driving management system 1 . Each process of the method of the embodiment of the present invention can be adjusted according to the implementation situation, and is not limited thereto.

FIG. 4 is a flow chart of a location evaluation method according to an embodiment of the present invention.

In step S410, the processor 16 of the vehicle-mounted device 10 obtains sensing data.

In one embodiment, the processor 16 of the vehicle-mounted device 10 obtains sensing data through one or more sensors 11 .

Specifically, as described above for the sensor 11 in FIG. 1 , depending on the type of the sensor 11 , the sensing data may be images, acceleration, speed, steering, angular velocity, vehicle speed, engine speed, and/or braking behavior. . However, the type of sensing data is not limited to the aforementioned types and may be related to any status or parameter of vehicle operation.

In step S420, the processor 16 of the vehicle-mounted device 10 determines the parking state according to the sensing data.

Specifically, the parking state represents whether the vehicle loaded, mounted, placed or built into the vehicle-mounted device 10 is parked. For example, the vehicle is turned off or stationary.

Regarding the determination of the parking status, in one embodiment, the sensing data includes positioning records and in-vehicle images. The positioning record is a record of one or more timestamps and the corresponding vehicle positions. The current position of the vehicle can be determined based on the positioning signal of the positioning device 12 and/or the sensing data of the inertial sensor. For example, when the positioning device 12 does not receive the positioning signal, the processor 16 can estimate the subsequent position based on the sensing data of the inertial sensor and through the inertial navigation technology. The in-vehicle images can be targeted at the driver and/or other passengers.

The processor 16 can determine the vehicle's stay time based on the positioning record. For example, the processor 16 confirms whether the position of the vehicle changes or changes within an allowable range (error related to the positioning device 12 or the sensor 11 ) at specific intervals (eg, 1 minute, 3 minutes, or 5 minutes). If the position does not change or the position change is within the allowable range, the processor 16 accumulates the dwell time.

The processor 16 of the vehicle-mounted device 10 can determine whether the passenger in the vehicle has left the seat based on the in-vehicle image. Depending on the task type, the identified passenger can be a passenger in the driver's seat. The processor 16 can apply known neural network or feature matching image recognition technology to determine whether the passenger has left the seat. For example, no person is detected in the driver's seat.

The processor 16 of the vehicle-mounted device 10 may determine the parking state according to the determination result of the stop time and the in-vehicle image. For example, if the passenger in the target seat (e.g., the driver's seat) leaves the seat and/or the stop time of the vehicle exceeds a certain time (e.g., 3 minutes, 5 minutes, or 10 minutes, and may be related to the statistical delivery time), the processor 16 may determine that the vehicle is in a parked state.

In one embodiment, the processor 16 of the vehicle-mounted device 10 can determine whether the current position of the vehicle is within a specific distance (e.g., 100, 300, or 500 meters) from the expected stop point to assist in confirming the stop state. In one embodiment, the processor 16 determines whether the vehicle is turned off based on the sensing data of the sensor 11. For example, the on-board diagnostic system receives a signal to stop the power supply to assist in confirming the stop state.

In one embodiment, the sensing data includes positioning records and external vehicle images. The external vehicle images may be images of the front or surrounding areas of the vehicle.

For example, FIG5 is a schematic diagram of a parking state according to an embodiment of the present invention. Referring to FIG5 , the external image is an image in front of the vehicle. In response to the parking state, the processor 16 may define the sensing data in the parking state according to the current position, the external image and the timestamp in the positioning record. For example, the current position, the external image and the timestamp are combined into a geostamp. The geostamp represents that the vehicle is parked at a specific location at a specific time and has an external image. The processor 16 may transmit the sensing data in the parking state to the server 30 via the communication transceiver 13.

In other embodiments, the vehicle-mounted device 10 may also transmit sensing data to the server 30 at a scheduled time or in response to a trigger condition, and may determine the parking status through the server 30 . That is, the server 30 executes step S420.

In step S430, the processor 36 of the server 30 obtains the parking location category corresponding to the sensing data in the parking state.

In one embodiment, the parking location categories include recommended parking categories, caution parking categories, and dangerous parking categories. It is recommended that parking categories comply with traffic regulations. "Recommended" means that the parking space is legal, such as roadside parking spaces, parking spaces in parking lots, open spaces at the docks of unloading warehouses, etc. Note that the parking category does not comply with traffic regulations but the corresponding accident risk (related to the occurrence rate of accidents) is less than the risk threshold. For example, "Caution" means that this location is not a normal parking location, but it is a temporary parking location. For example, yellow line or open space. The dangerous parking category does not comply with traffic regulations and the corresponding accident risk is not less than the risk threshold. "Danger" refers to parking locations that are illegal and pose a risk of causing an accident. For example, the red line area. Does Chinese mean acceleration of gravity? XD

In one embodiment, the parking location categories also include other/not recommended parking categories. For example, "Other/Not Recommended" is a location the system doesn't recognize, or a one-way street without yellow or red lines. However, there may be other variations to the parking location categories and may be changed based on the user's needs.

In one embodiment, the processor 36 of the server 30 can determine the initial category corresponding to the sensed data according to the classification rules. Classification rules relate to parking legality. Classification rules can refer to the government's public database of legal parking locations, road regulations and parking location data accumulated by drivers of each fleet or drivers in a specific area. The initial category is one of those parking location categories. That is, the processor 36 of the server 30 determines which of the parking position categories the sensing data in the parking state belongs to according to the classification rules.

For example, for the recommended parking category: this location is fixed as a driver parking location with a certain percentage (e.g., 25, 30, or 50%), the external image detects parking space features, or other vehicles in the external image are parked for more than a specific time (e.g., 3, 5, or 10 minutes), and this location is within a specific distance (e.g., 50, 100, or 200 meters) from the delivery location.

Regarding the parking categories to be noted: the road or roadside features detected by the vehicle's external image are permitted for temporary short-term parking within the country's traffic regulations; or, there is no previous record of other drivers parking at this location, this location is not a parking space, and this location is a certain distance away from the delivery location (for example, more than 300 meters); the vehicle is parked in a parking space, but the distance to the delivery location exceeds a certain range (for example, more than 500 meters); this location is an open space, there is no record of other drivers parking, and it is a certain distance away from the delivery location (for example, more than 200 meters); this location is more than a certain time away from the parking during the delivery, for example, the vehicle has not encountered any other vehicles on the road within an hour.

For the dangerous parking category: There has been a record of drivers parking at this location, but the external image detection does not show any road features that can be parked. For example, there are no parking spaces and no other vehicles parked nearby for more than a certain period of time (e.g., 1, 3, or 5 minutes); the location is a one-way street but the road is significantly larger than an ordinary road (e.g., more than 1.5 times the width of the road).

For other/not recommended parking categories: This location cannot be identified as parking (for example, it does not belong to the aforementioned categories); it is more than a certain distance from the delivery location (for example, more than 800 meters); this location has been compared with map information and The road features in the image outside the vehicle are then determined to be a no-parking location (for example, a sidewalk).

In one embodiment, the vehicle's route is determined. A route is a navigation path for one or more stop points by the navigation system or a path edited by input operations. Stopping points are, for example, delivery locations, attractions or restaurants. The processor 16 may obtain street view images within the evaluation range of the stop point (for example, a radius of 300 meters, 500 meters, or 1 kilometer). For example, the street view image is downloaded through the communication transceiver 13 . Then, the processor 16 can determine the parking space or vacant space through image recognition technology based on neural network or feature comparison, or transmit it to the server 30 through the communication transceiver 13 and use the server 30 to identify the parking space, vacant space or other parking spaces. Bit. These parking spaces can be combined with their location information and reviewed later.

In one embodiment, parking spaces, open spaces or other parking spaces may be obtained by receiving the user's recognition operation on the navigation map/street view image on the viewer through an input device (not shown). These parking spaces can be combined with their location information and reviewed later.

The processor 36 of the server 30 can obtain the audit result of the initial category. The audit result may be obtained by receiving a mark operation on the viewer for the initial category by a user (eg, a fleet manager or a data processor) through an input device (not shown). The marking operation can be to confirm the initial category or to modify the initial category to another parking location category.

For example, FIG. 6 is a schematic diagram of parking location categories according to an embodiment of the present invention. Referring to FIG. 6 , the viewer (implemented through the display of the vehicle-mounted device 10 , the server 30 or the remote device) may display the parking location category according to the initial category (for example, the parking location category may be the advice category PT1 , the attention category PT2 and/or the hazard category PT3 Parking location) classified images (for example, images outside the car). For example, the images IM11~IM14, IM21, IM31, and IM32 of parking locations belonging to the recommendation category PT1, the attention category PT2, and the hazard category PT3 are respectively located in different columns in the user interface displayed in the viewer, and the user interface uses the category name ("Advice", "Caution", "Danger" as shown in the picture) distinguish different types of images. It is recommended that the contents of images IM11~IM14 of category PT1 may be vehicles located in vehicle grids, unmarked areas on the roadside, or in open spaces. Note that the content of image IM21 of category PT2 may be that the vehicle is located in a pause area or a designated parking area. The content of images IM31 and IM32 in the risk category PT3 may be that the vehicle is located on a no-stop line, sidewalk or bus parking area.

It should be noted that in other embodiments, the processor 36 may directly use the initial category as the parking location category without review, or determine the parking location category entirely by marking operations.

In step S440, the processor 36 of the server 30 trains the location suggestion model based on the parking location category and the sensing data.

Specifically, the location recommendation model is used to recommend parking locations, and the location recommendation model is trained by a machine learning algorithm. The machine learning model/algorithm (for example, applying a known neural network, support vector machine (SVM), or random forest) can analyze training samples to obtain rules from them, thereby predicting unknown data through the rules. The location recommendation model is the machine learning model constructed after learning, and is used to infer the parking location category or recommended parking location based on evaluation data (for example, sensor data). The parking spots are divided into suitable parking spots, temporary parking spots or high-risk parking spots (for example, recommended, cautionary or dangerous parking spots). Through big data analysis and image recognition results, the database of suitable parking spots is automatically summarized and gradually established, so as to make the most recommended parking spots for novice drivers in the future. It is worth noting that since the parking spot is a regional concept and not a fixed location, it is necessary to gradually establish the recommended parking area and range through data analysis and learning.

In addition to suggestions for parking locations, embodiments of the present invention can also evaluate (prone) violation locations.

Fig. 7 is a flow chart of determining a violation location according to an embodiment of the present invention. Referring to Fig. 7, in step S710, the processor 36 of the server 30 may count the number of violation events.

In one embodiment, the sensing data includes a positioning record, and the location of the violation event is within a statistical range (e.g., a radius of 10, 50, or 100 meters) from the location of the positioning record. For example, the violation event is obtained from a violation record reported by the driver himself or provided by the operator himself, obtained by connecting the communication transceiver 33 to a government public database (e.g., recording the time, location, number of casualties, and longitude and latitude), or obtained by image recognition of the vehicle-mounted device 10 or judging driving misconduct based on inertial sensing data (e.g., distracted driving, fatigue driving, driving in the wrong direction, running a red light, sudden braking, sharp turns, idling, or speeding). The statistical time period can be 3 days, one week or one month.

In step S720, the processor 36 of the server 30 may determine whether the event frequency exceeds a frequency threshold (eg, 10, 20, or 30 times).

In step S730, for the violation events that exceed the frequency threshold, in response to the event frequency exceeding the frequency threshold, the processor 36 of the server 30 may classify the event locations corresponding to the violation events.

For example, the violation events corresponding to the event locations within a radius of 10 or 20 meters are classified into the same event group.

In step S740, the processor 36 of the server 30 may define the violation location according to the event location. For example, the area corresponding to a certain event group is marked as a (probable) violation location. Other events that do not exceed the number threshold value may be defined as violation location candidates.

Regarding the parking location prompt, in one embodiment, the processor 36 of the server 30 can determine the first parking location of the recommended parking category corresponding to the stop point in the route through the location suggestion model. A route is a navigation path for one or more stop points by the navigation system or a path edited by input operations. Stopping points are, for example, delivery locations, attractions or restaurants.

In one embodiment, the processor 36 of the server 30 can determine whether the street view image within the recommended range of the stop point (for example, a radius of 100, 200 or 300 meters) is a recommended parking category through the location suggestion model. If it belongs to the recommended parking category, the processor 36 uses the location corresponding to the street view image as the first parking location of the recommended parking category. Alternatively, these first parking locations have been stored in the parking database, and the processor 36 of the server 30 can search for the first parking locations within the recommended range from the parking database.

In one embodiment, if there are a plurality of first parking spots, the processor 36 of the server 30 may sort these first parking spots according to the number of previous parking times. For example, the first parking spot with a higher number of parking times is recommended first.

For example, FIG. 8 is a schematic diagram of parking location suggestions according to an embodiment of the present invention. Please refer to Figure 8. The route includes stopping points P1~P5. The processor 36 of the server 30 can determine the parking location belonging to the recommended category PT1 within the recommended range (for example, 200 meters) with the stay point P1 as the center of the circle, and within the recommended range (for example, 200 meters) with the stay point P2 as the center. Determine the parking location that belongs to the caution category PT2 within 50 feet), and determine the parking location that belongs to the hazard category PT3 within the recommended range (for example, 200 meters) with the stop point P4 as the center. Different parking location categories can be presented with different visual patterns, colors, sizes or text.

In one embodiment, in response to the vehicle being within the recommended range of the parking spot, the processor 16 of the vehicle-mounted device 10 may provide the first parking spot and its street view image through the display 14 or the passenger's remote device 40 .

For example, FIG. 9A is a schematic diagram of parking location suggestions of the remote device 40 according to an embodiment of the present invention. Referring to FIG. 9A , when the vehicle is within 200 meters from the parking point, the remote device 40 can display the navigation information I1 of the parking location P6 belonging to the recommended parking category (for example, the estimated arrival time is 1 minute and the distance is 160 meters). ) and street view image SV1.

As another example, FIG. 9B is a schematic diagram of parking location suggestions of the remote device 40 according to an embodiment of the present invention. Referring to FIG. 9B , the vehicle continues to move, and the remote device 40 can display the navigation information I2 of another parking location P7 belonging to the recommended parking category (for example, the estimated arrival time is 3 minutes and the distance is 320 meters) and the street view image SV2.

In addition to the parking location category suggestions, the remote device 40 or the vehicle-mounted device 10 may also provide suggestions for illegal locations. For example, when the location information obtained by the positioning device 12 indicates that the vehicle moves within a warning range (e.g., a radius of 50, 100, or 300 meters) from the illegal location, the remote device 40 or the vehicle-mounted device 10 provides the illegal location and its street view image.

In one embodiment, the processor 36 of the server 30 can obtain the parking location corresponding to the street view image within the evaluation range (for example, 100, 200 or 300 meters) of the stop point in the route through the location suggestion model or based on the audit results. Category and provide secondary parking locations determined to be the recommended parking category from Street View imagery within the assessment area. That is to say, the processor 36 of the server 30 can determine whether the street view image within the evaluation range of the stay point is a recommended parking category through the location suggestion model. If it belongs to the recommended parking category, the processor 36 of the server 30 uses the location corresponding to the street view image as the second parking location of the recommended parking category. This second parking location is stored in the parking database for subsequent access.

In one embodiment, the processor 16 of the vehicle-mounted device 10 may provide task content through the display 14 or the passenger's remote device 40 .

FIG. 10A is a schematic diagram of a task preview of a remote device according to an embodiment of the present invention. Please refer to Figure 10A, taking delivery as an example. Task content I3 includes a total of several trips to be made, and several delivery points for each trip. The weather of the day is displayed to remind the driver whether the distance is increased or decreased due to weather effects.

FIG. 10B is a schematic diagram of a task preview of a remote device according to an embodiment of the present invention. Referring to FIG. 10B , the task content I4 includes the delivery details of each task (for example, consignee, address, driving distance, and easy violation event reminder). Each task is also equipped with a one-touch call function to facilitate users to ask the consignee in advance whether it is convenient to receive the goods at home. If the consignee is absent, the remote device 40 can skip the order and automatically re-plan the route by receiving a click operation on the skip button in the user interface.

In addition, the remote device 40 can also display the recommended stopover candidate locations for this mission in advance as shown in FIG. 9A and FIG. 9B , and display the parking hotspot map location and street view images. This allows the driver to directly preview the route through the map mode, and can intuitively observe the delivery points, the warning icons of possible violations, and the recommended parking hotspot locations.

In addition, the driver can use the functions provided by the vehicle-mounted device 10 during the delivery process. For example, navigation delivery routes, reminders of possible violations (voice and screen), parking hotspot suggestions, and delivery detail viewing (for example, consignee, address, distance, one-click call, skip task). During the delivery process, parking locations or violation locations can be reminded by voice, but voice reminders can also be turned off. The switching of the voice reminder function can also be notified to the remote device 40 of the management end to facilitate understanding of the reminder status.

On the other hand, in order to achieve energy conservation and carbon reduction, help fleet managers reduce cost problems caused by unnecessary energy consumption, and implement corporate responsibility for environmental protection, it is necessary to further analyze whether driving behavior meets energy conservation requirements.

FIG11 is a flow chart of a method for analyzing driving behavior according to an embodiment of the present invention. Referring to FIG11 , the description of step S111 may refer to step 410 of FIG4 , which will not be described in detail here.

The processor 36 of the server 30 may determine the energy saving score of the sensed data based on one or more energy saving factors (step S112). Specifically, the energy saving factor is a factor that affects the energy consumption of the vehicle.

In one embodiment, the sensing data includes the vehicle speed, the energy saving factor includes cruise control, and the energy saving score includes a sub-score of the cruise control.

Figure 12 is a schematic diagram of cruise control analysis according to an embodiment of the present invention. Referring to FIG. 12 , the processor 36 of the server 30 can count the first accumulated difference between the (actual) vehicle speed 121 and the energy-saving cruising speed 122 during the journey. For example, in units of seconds, compare the absolute value of the difference between the energy-saving cruising speed 122 minus the vehicle speed 111, and accumulate the absolute value of the difference for each second.

In one embodiment, the trip may be the total trip of a day or trips within another time range. When the vehicle is completely stationary, no difference is accumulated. The eco-cruise speed may vary depending on the type of vehicle and/or road, and is related to the driving speed in eco-mode.

On the other hand, the processor 36 of the server 30 can count the second accumulated difference between the (actual) vehicle speed 121 and the road speed limit 123 during the journey. For example, in seconds, the absolute value of the difference between the road speed limit 123 and the vehicle speed 111 (as shown in the bottom part of the figure) is compared, and the absolute value of the difference is accumulated every second. When the vehicle is completely stationary, the difference is not accumulated.

Then, the processor 36 of the server 30 may determine the sub-score of the cruise control according to the first accumulated difference and the second accumulated difference. When the energy-saving cruise speed 122 is greater than the road speed limit 123, the processor 36 of the server 30 may only accumulate the first accumulated difference. When the road speed limit 123 is greater than the energy-saving cruise speed 122, the processor 36 of the server 30 may only accumulate the second accumulated difference.

Next, the processor 36 of the server 30 adds up the first cumulative difference and the second cumulative difference. If the summed value accounts for a lower proportion of the total time of the journey, the cruise control sub-score is higher. That is, the higher the proportion of the actual driving speed 121 close to the energy-saving cruising speed 122 or the road speed limit 123, the more energy is saved. If the summed value accounts for a higher proportion of the total time of the journey, the cruise control sub-score is lower. That is, the lower the proportion of the actual driving speed 121 close to the energy-saving cruising speed 122 or the road speed limit 123, the more energy is consumed.

In one embodiment, the sensing data includes the braking behavior of the vehicle, the energy saving factor includes coasting, and the energy saving score includes a sub-score of coasting. Figure 13 is a schematic diagram of sliding analysis according to an embodiment of the present invention.

Please refer to Figure 13, the definition of coasting can be that the vehicle is 50 meters before the stop line before the red light or 50 meters away from the vehicle in front (but the distance is not limited to this), no vehicle can be detected in the image outside the vehicle, the vehicle has released the accelerator and the speed is greater than zero. The processor 36 can count the number of braking times of coasting to one or more stopping points in the journey. For example, before reaching the first stopping point, the number of braking times is 3 times. The processor 36 of the server 30 can determine the number of coasting times in the journey based on the number of braking times. For example, the ratio of all braking times in the journey to the number of stopping points. If this braking ratio is higher, the number of coasting times is lower and more energy is consumed. If this braking ratio is lower, the number of coasting times is higher and more energy is saved.

In one embodiment, the sensing data includes speeding behavior of the vehicle, the energy saving factor includes speeding, and the energy saving score includes a sub-score of speeding. The processor 36 of the server 30 can combine multiple over-speed actions within one second into one action. If the speed difference between the maximum vehicle speed and the speed limit is greater than the speeding threshold (for example, 5 kilometers, 10 kilometers or 15 kilometers per hour), the processor 36 of the server 30 defines the speeding behavior. The processor 36 of the server 30 may determine the overspeed event based on the duration of the overspeed behavior. If the time (i.e., duration) defined as the interval between two overspeeding behaviors exceeds an interval threshold (eg, 5 seconds, 10 seconds, or 20 seconds), the processor 36 of the server 30 defines the overspeeding behavior for this duration. for a speeding incident. If the speed difference is not greater than the overspeed threshold or the duration does not exceed the interval threshold, the processor 36 of the server 30 continues to detect subsequent overspeed behaviors.

The processor 36 of the server 30 can count the number of speeding incidents during the journey, and determine the speeding score based on the number of speeding incidents. If the number of overspeeding is greater (for example, it is greater than the corresponding number threshold), the overspeeding score will be lower and the energy consumption will be more. If the number of overspeeding is lower (for example, less than the corresponding number of times threshold), the number of overspeeding times will be higher and more energy will be saved.

In one embodiment, the sensing data includes vehicle speed and engine speed, the energy saving factor includes idle speed, and the energy saving score includes a sub-score of idle speed. Similarly, the processor 36 of the server 30 can combine multiple idle actions within one second into one action. The processor 36 may determine the idling behavior based on vehicle speed and engine speed. If the statistical vehicle speed (for example, the average vehicle speed or the median vehicle speed) is zero but the engine speed is greater than zero, the processor 36 of the server 30 defines an idling behavior. If the time (i.e., duration) defined as the interval between two idle behaviors exceeds a cumulative threshold (eg, 10 seconds, 20 seconds, or 1 minute), the processor 36 of the server 30 defines the idle behavior for this duration. for idling events. That is, the idling event occurs when the vehicle speed is zero but the engine speed is greater than zero.

The processor 36 of the server 30 may count the accumulated idle time of idle events during the journey. When an idle event is determined, the processor 36 of the server 30 continues to count the duration of the idle event as the accumulated idle time. The processor 36 of the server 30 may determine the idle sub-score based on the accumulated idle time. The cumulative idling time is, for example, no more than 3 minutes (determined according to regulations). If the accumulated idle time is longer (for example, greater than the corresponding time threshold), the idle speed score will be lower and the energy consumption will be greater. If the accumulated idle time is lower (for example, less than the corresponding time threshold), the idle sub-score will be higher and more energy will be saved.

In one embodiment, the sensing data includes acceleration and speed of the vehicle, the energy saving factors include expected events, and the energy saving score includes sub-scores of expected events. The processor 36 of the server 30 can determine the emergency braking event based on acceleration and speed. A sudden braking event occurs when the acceleration in the direction of travel of the vehicle is lower than the lower limit of acceleration and the change period of the vehicle speed is greater than the upper limit of change.

For example, the minimum value of the linear accelerometer X axis (corresponding to the direction of travel) is less than -0.28g (i.e., the lower limit of acceleration, symbol g is the unit of gravity), the average accumulated amount of the linear accelerometer X axis during this period (in seconds) is less than or equal to -2.8 seconds, and the average change period of the vehicle speed during this period (in seconds) is greater than or equal to 5.0 seconds (i.e., the upper limit of change). The occurrence of sudden braking events may be caused by the vehicle in front or an accident, so the detection of sudden braking events can be used to predict events.

The processor 36 of the server 30 can combine multiple sudden braking behaviors within one second into one behavior. In the statistical process, if the average linear accelerometer X-axis is less than, for example, -0.17g and the average vehicle speed is greater than zero, the processor 36 of the server 30 can define it as a sudden braking candidate behavior.

If the interval (i.e., duration) between two emergency candidate actions exceeds the interval threshold (for example, 5 seconds, 10 seconds or 20 seconds), the processor 36 of the server 30 will delete the emergency candidate during this duration. The behavior is defined as urgent behavior. Otherwise, the processor 36 of the server 30 continues to detect subsequent emergency candidate behaviors.

In some embodiments, the processor 36 of the server 30 can also determine the speed and acceleration of the traffic ahead by identifying the image outside the vehicle, and use it as an assessment of the sudden braking event. Alternatively, the processor 36 can also identify the distance to the preceding vehicle in the image outside the vehicle, and prompt the remote device 40 or the vehicle-mounted device 10 of the distance to the preceding vehicle, so as to reduce sudden braking behavior or other abnormal braking behavior.

The processor 36 of the server 30 can count the number of emergency braking events during the journey and determine the score of the expected event according to the number of emergency braking events. If the number of emergency braking events is greater (e.g., greater than the corresponding number threshold), the number of expected events is lower and more energy is consumed. If the number of emergency braking events is less (e.g., less than the corresponding number threshold), the number of expected events is higher and more energy is saved.

In one embodiment, the sensing data includes the engine speed of the vehicle, the energy-saving factor includes the green band, and the energy-saving score includes the sub-score of the green band. The processor 36 can count the cumulative effective speed time during the journey when the engine speed meets the target speed (e.g., 1400-1600 revolutions per minute). The target speed is determined based on the vehicle model and road conditions and is related to the engine speed in the energy-saving mode. For example, the target speed is lower on a downhill section, but higher on an uphill section. When the engine speed is within the allowable range of the target speed (e.g., 30-50 revolutions per minute), the processor 36 of the server 30 can count and obtain the cumulative effective speed time. The processor 36 can determine the sub-score of the green band based on the cumulative effective speed time. If the accumulated effective speed time accounts for more of the total time of the journey (for example, greater than the corresponding time threshold), the expected event score is lower and more energy is consumed. If the accumulated effective speed time accounts for less of the total time of the journey (for example, less than the corresponding time threshold), the expected event score is higher and more energy is saved.

In one embodiment, the sensing data includes the vehicle interior temperature, the energy saving factor includes the use of air conditioning, and the energy saving score includes the number of times the air conditioning is used. The processor 36 of the server 30 can calculate the accumulated energy saving time during the journey when the vehicle interior temperature meets the target temperature. The target temperature is determined based on the climate.

For example, summer corresponds to the lower temperature limit, and winter corresponds to the upper temperature limit. When the temperature inside the vehicle is within the allowable range of the target temperature (for example, 1 degree or 3 degrees), the processor 36 of the server 30 can time the timing and obtain the accumulated energy saving time.

In one embodiment, the processor 36 of the server 30 may determine the number of times the air conditioner is used based on the accumulated energy-saving time. If the accumulated energy-saving time accounts for more of the total time of the journey (e.g., greater than the corresponding time threshold), the number of times the air conditioner is used is lower and more energy is consumed. If the accumulated energy-saving time accounts for less of the total time of the journey (e.g., less than the corresponding time threshold), the number of times the air conditioner is used is higher and more energy is saved.

For example, for each temperature and each specific time (e.g., one minute as one unit), the temperature inside the car is lower than the standard value range in summer and is considered as uneconomical driving, while the temperature inside the car is higher than the standard value range in winter. The score is calculated based on the proportion of each unit in the unhealthy driving range to the entire driving time (i.e., the total time of the journey).

Table (1) is an example to illustrate the sub-scores of multiple energy saving factors. Table (1) Energy saving factors Statistical results Times score Cruise control 80% meets the calculation criteria 80 Coasting (any gear) 58% meets the calculation criteria Section A: Braking 2 times 50% Section B: Braking 3 times 33% Section C: Braking 2 times 100% Section D: Braking 2 times 50% 58 Over speed The fleet had 26 speeding incidents in total, and driver A had 17 speeding incidents, which reached 65% compared to the average. 35 Idling The accumulated idling time for a single trip is 90 seconds, which is only 66% of the standard value. 66 Anticipation There were 13 accidents ahead and the driver had to brake suddenly 6 times (46%), which is only 39% of the standard value. 39 Green band 76% met the calculation criteria 76 Air conditioning use 94% met the calculation criteria 94

In some embodiments, the sensed data may also be tire pressure or regular maintenance notifications. The processor 36 of the server 30 may notify the remote device 40 or the vehicle-mounted device 10 through the communication transceiver 33 based on the detection result that the tire pressure is too low (for example, less than 10% of the standard) or an insurance notification.

In one embodiment, there are multiple energy saving factors. The processor 36 of the server 30 can assign weights to these energy-saving factors, and perform weighting operations and/or normalization operations on the sub-scores of these energy-saving factors according to the corresponding weights to obtain energy-saving scores.

Referring to FIG. 11 , the processor 36 of the server 30 generates a driving behavior report based on the energy saving score (step S113 ). Specifically, driving behavior reports illustrate the strengths and weaknesses of energy saving scores. The processor 36 of the server 30 can select a reference score from multiple energy-saving scores (for example, for different vehicles or drivers), and compare these energy-saving scores with the reference scores respectively, so as to know whether the driving behavior is energy-saving, and then serve as a guide for energy-saving driving. Reference learning basis.

In one embodiment, the sensing data corresponds to multiple sections in the route, and these sections are classified according to road characteristics. For example, a straight section, a turning section, an uphill section, a downhill section, a highway section, or a non-road section. The processor 36 of the server 30 can generate energy-saving scores for these sections respectively, and compare the energy-saving scores of these sections with the corresponding reference scores respectively. The driving behavior report includes the comparison results of these sections. In this way, by comparing different driving behaviors for each section, it can help the driver understand how to drive in the most fuel-efficient way under the same road conditions, and then serve as an energy-saving driving indicator.

In one embodiment, the processor 36 of the server 30 may determine the driving status based on the driving behavior report. For example, the processor 36 of the server 30 grades the score difference between the energy saving score and the reference score, and evaluates the driving state based on the grades. The higher the level, the better the driving condition. The lower the level, the worse the driving condition.

The processor 36 of the server 30 can notify the remote device 40 or the vehicle-mounted device 10 through the communication transceiver 33 that the driving status meets the notification condition. The notification condition is related to the ranking of the energy saving score.

For example, the processor 36 of the server 30 is ranked according to the level. The notification condition is the lower ranked vehicle or driver. Therefore, the processor 36 of the server 30 can remind the remote device 40 or the vehicle-mounted device 10 of the driving status through the communication transceiver 33 .

For example, FIG. 14 is a schematic diagram of driver analysis according to an embodiment of the present invention. Referring to Figure 14, assuming there are three series, the energy saving score of series D1 is closest to the reference score, the energy saving score of series D2 is the second closest to the reference score, and the energy saving score of series D3 is farthest from the reference score. The administrator's remote device 40 can display the location of each vehicle on the electronic map, and use different visual patterns/colors/texts to distinguish vehicles of different levels. For level D3, the user interface also provides the option of direct call so that the manager can remind the driver by voice.

In one embodiment, the processor 36 of the server 30 can count the violations of each driver. Violations include distracted driving, driving while fatigued, driving in the wrong direction, running red lights, sudden braking, sharp turns, idling or speeding.

In one embodiment, during the journey, if the number of violations of the violation behavior exceeds the corresponding number threshold, the processor 36 of the server 30 can notify the remote device 40 or the vehicle-mounted device 10. For example, the manager's remote device 40 prominently reminds the vehicle with too many violations. In addition, the user interface can also provide a direct call option so that the manager can remind the driver by voice.

In one embodiment, the processor 36 of the server 30 can identify whether the passenger in the driver's seat in the in-car image is a registered driver. For non-registered drivers, the user interface also provides a direct call option so that managers can remind the driver via voice.

FIG15 is a schematic diagram of a notification of a remote device 40 according to an embodiment of the present invention. Referring to FIG15 , the driver's remote device 40 presents notification content I5. For example, too many violations, too low energy-saving score, or not the driver, please pay attention.

In summary, in the location evaluation method, driving behavior analysis method and driving management system of the embodiments of the present invention, sensing data is collected, feature values are found from the sensing data, and combined with the review to serve as training data for the algorithm, so as to find recommended locations suitable for parking. Sensors and image capture devices on IoT devices or transportation vehicles can continuously collect sensing data, and the system continuously trains the algorithm for the required recommended locations or updates the database. The embodiments of the present invention can provide an interface for data scientists or managers to modify, review and annotate the collected sensing data. The embodiments of the present invention also provide an interface for developers to adjust algorithm parameters to ensure that the recommended locations meet the user's expectations for the recommended locations. The embodiment of the present invention also provides a scoring mechanism, which automatically compares driving behaviors according to route segments, finds the driving behaviors of drivers with the highest energy-saving scores and uses them as the standard basis. The comparison of road segments (benchmark) can be used as a basis for evaluating driving behaviors, and a driving behavior report can be generated, which can be used as a reference for driving training. In this way, the problem of low efficiency caused by novice drivers' unfamiliarity with road conditions can be improved, helping to improve driving safety, saving the operating costs caused by traditional driver training and energy waste due to bad driving habits, and improving the efficiency of novice drivers in performing tasks.

Although the present invention has been disclosed above through embodiments, they are not intended to limit the present invention. Anyone with ordinary knowledge in the technical field may make some modifications and modifications without departing from the spirit and scope of the present invention. Therefore, The protection scope of the present invention shall be determined by the appended patent application scope.

1: Driving management system 10: Vehicle-mounted device 20:Network access device 30:Server 40:Remote device 11: Sensor 12: Positioning device 13: Communication transceiver 14:Display 15:Memory 16: Processor 33: Communication transceiver 35:Memory 36: Processor S410~S440, S710~S740, S111~S113: steps PT1: Create categories PT2: Pay attention to categories PT3: Hazard Category IM11~IM14, IM21, IM31, IM32: image P1~P5: stop point I1, I2: Navigation information SV1, SV2: street view images I3, I4: task content 121:Vehicle speed 122: Energy-saving cruising speed 123:Road speed limit D1~D3: series I5: Notification content

FIG. 1 is a schematic diagram of a driving management system according to an embodiment of the present invention. FIG. 2 is a component block diagram of a vehicle-mounted device according to an embodiment of the present invention. FIG. 3 is a component diagram of a server according to an embodiment of the present invention. FIG. 4 is a flow chart of a location evaluation method according to an embodiment of the present invention. Figure 5 is a schematic diagram of a parking state according to an embodiment of the present invention. Figure 6 is a schematic diagram of parking location categories according to an embodiment of the present invention. Figure 7 is a flow chart of violation location determination according to an embodiment of the present invention. FIG. 8 is a schematic diagram of parking location suggestions according to an embodiment of the present invention. FIG. 9A is a schematic diagram of parking location suggestions of a remote device according to an embodiment of the present invention. FIG. 9B is a schematic diagram of parking location suggestions of a remote device according to an embodiment of the present invention. FIG. 10A is a schematic diagram of a task preview of a remote device according to an embodiment of the present invention. FIG. 10B is a schematic diagram of a task preview of a remote device according to an embodiment of the present invention. FIG. 11 is a flow chart of a driving behavior analysis method according to an embodiment of the present invention. Figure 12 is a schematic diagram of cruise control analysis according to an embodiment of the present invention. Figure 13 is a schematic diagram of sliding analysis according to an embodiment of the present invention. Figure 14 is a schematic diagram of driver analysis according to an embodiment of the present invention. FIG. 15 is a schematic diagram of notification by a remote device according to an embodiment of the present invention.

S410~S440: Steps

Claims (20)

A location assessment method that includes: Obtain sensing data; Determine a parking status based on the sensing data; Obtain a parking location category corresponding to the sensing data in the parking state; and A location recommendation model is trained according to the parking location category and the sensing data, wherein the location recommendation model is used to recommend a parking location. The position evaluation method as described in claim 1, wherein the sensing data includes a positioning record and an in-car image, and the steps of determining the parking status based on the sensing data include: Determine a stay time of a vehicle based on the positioning record; Determine whether a passenger in the vehicle leaves his seat based on the in-vehicle image; and The parking status is determined based on the dwell time and the decision result of the image in the car; The parking status represents whether the vehicle is parked. The position evaluation method as described in claim 1, wherein the sensing data includes a positioning record and an external image of the vehicle, and the evaluation method further includes: In response to the parking state, the sensing data in the parking state is defined based on a current location in the positioning record, the exterior image and a timestamp. The location assessment method as described in claim 1, wherein the parking location category includes a recommended parking category, and the assessment method further includes: Determine a first parking location of the proposed parking type corresponding to a stop point in a route through the location suggestion model; and In response to the fact that the vehicle is located within a recommended range of the stopping point, the first parking location and its street view image are provided. The location evaluation method as described in request 1 further includes: Obtain street view images within an evaluation range of a stop point on a route; Obtain the parking location category corresponding to the street view image within the assessment range; and A secondary parking location determined to be the proposed parking category is provided from street view imagery within the assessment area. The method for evaluating a location as described in claim 1, wherein the step of obtaining the parking location category corresponding to the sensing data in the parking state includes: Determining an initial category corresponding to the sensing data according to a classification rule, wherein the classification rule is related to parking legality; Obtaining a review result of the initial category; and Determining the parking location category according to the review result. The location evaluation method as described in request 1 further includes: Counting an event number of a violation event, wherein the sensing data includes a positioning record, and the location where the violation event occurs is within a statistical range from a location in the positioning record; Determine whether the number of events exceeds the once threshold; In response to the number of incidents exceeding the threshold, classify at least one incident location corresponding to the violation incident; and A violation location is defined based on the at least one event location. A method for evaluating a location as described in claim 1, wherein the parking location category includes a recommended parking category, a caution parking category, and a dangerous parking category, the recommended parking category complies with a traffic regulation, the caution parking category does not comply with the traffic regulation but the corresponding accident risk is less than a risk threshold value, and the dangerous parking category does not comply with the traffic regulation and the corresponding accident risk is not less than the risk threshold value. A method for analyzing driving behavior, comprising: Obtaining a sensing data; Determining an energy-saving score of the sensing data based on at least one energy-saving factor, wherein the at least one energy-saving factor is a factor that affects the energy consumption of a vehicle; and Generating a driving behavior report based on the energy-saving score, wherein the driving behavior report describes the pros and cons of the energy-saving score. The driving behavior analysis method as described in claim 9, wherein the sensing data corresponds to multiple road sections in a route, the road sections are classified according to road characteristics, and the step of generating the driving behavior report based on the energy saving score includes : Generate energy saving scores for these road sections respectively; and The energy saving scores of the road sections are compared with the corresponding reference scores respectively, where the driving behavior report includes the comparison results of the road sections. The driving behavior analysis method as described in claim 9, wherein the sensing data includes a speed of the vehicle, the at least energy saving factor includes a certain speed cruise, the energy saving score includes a score of the fixed speed cruise, and based on the The steps of determining the energy-saving score of the sensing data by at least one energy-saving factor include: Calculate a first cumulative difference between the vehicle speed and the energy-saving cruising speed during a journey; Calculate a second cumulative difference between the vehicle speed and a road speed limit during the journey; and The score of the cruise control is determined based on the first cumulative difference and the second cumulative difference. A method for analyzing driving behavior as described in claim 9, wherein the sensing data includes a braking behavior of the vehicle, the at least one energy-saving factor includes a coasting, the energy-saving score includes a score of the coasting, and the step of determining the energy-saving score of the sensing data based on the at least one energy-saving factor includes: Counting a number of braking times of the braking behavior at at least one stop point in a journey; and Determining the score of the coasting based on the number of braking times. A method for analyzing driving behavior as described in claim 9, wherein the sensing data includes a speeding behavior of the vehicle, the at least one energy-saving factor includes an overspeeding, the energy-saving score includes a score of the overspeeding, and the step of determining the energy-saving score of the sensing data based on the at least one energy-saving factor includes: Determining a speeding event based on a duration of the speeding behavior; Counting the number of speeding times of the speeding event in a journey; and Determining the score of the speeding event based on the number of speeding times. A method for analyzing driving behavior as described in claim 9, wherein the sensing data includes a vehicle speed and an engine speed of the vehicle, the at least one energy-saving factor includes an idle speed, the energy-saving score includes a sub-score of the idle speed, and the step of determining the energy-saving score of the sensing data based on the at least one energy-saving factor includes: Determining an idle event based on the vehicle speed and the engine speed, wherein the idle event occurs when the vehicle speed is zero but the engine speed is greater than zero; Counting a cumulative idle time of the idle event in a journey; and Determining the sub-score of the idle speed based on the cumulative idle time. A method for analyzing driving behavior as described in claim 9, wherein the sensing data includes an acceleration and a vehicle speed of the vehicle, the at least one energy-saving factor includes an expected event, the energy-saving score includes a score of the expected event, and the step of determining the energy-saving score of the sensing data based on the at least one energy-saving factor includes: Determining an emergency braking event based on the acceleration and the speed, wherein the emergency braking event occurs when the acceleration in the direction of travel of the vehicle is lower than an acceleration lower limit and the change period of the vehicle speed is greater than an upper change limit; Counting the number of emergency braking events of the emergency braking event in a journey; and Determining the score of the expected event based on the number of emergency braking events. The driving behavior analysis method as described in claim 9, wherein the sensing data includes an engine speed of the vehicle, the at least energy-saving factor includes a green band, and the energy-saving score includes a score of the green band. , and the step of determining the energy-saving score of the sensing data based on the at least energy-saving factor includes: Calculate a cumulative effective speed time during a journey when the engine speed matches a target speed, where the target speed is determined based on the vehicle model and road conditions; and The score of the green belt is determined based on the accumulated effective speed time. The driving behavior analysis method as described in claim 9, wherein the sensing data includes an interior temperature of the vehicle, the at least energy-saving factor includes an air-conditioning usage, the energy-saving score includes a score of the air-conditioning usage, and based on The step of determining the energy-saving score of the sensing data by the at least energy-saving factor includes: Calculate a cumulative energy saving time during a journey when the temperature inside the car meets a target temperature, where the target temperature is determined based on the climate; and The fraction used by the air conditioner is determined based on the accumulated energy saving time. The driving behavior analysis method as described in claim 9 further includes: Determining a driving state based on the driving behavior report; and Notifying that the driving state meets a notification condition, wherein the notification condition is related to the ranking of the energy saving score. A driving management system includes: A server, which is communicatively connected to a vehicle-mounted device, wherein the server obtains a parking location category corresponding to a sensing data of the vehicle-mounted device in a parking state, and trains a location recommendation model based on the parking location category and the sensing data, and the location recommendation model recommends a parking location. A driving management system includes: A server, communicatively connected to a vehicle-mounted system, wherein the server determines an energy-saving score of a sensing data of the vehicle-mounted device according to at least one energy-saving factor, and generates a driving behavior report according to the energy-saving score, wherein the at least one energy-saving factor is a factor affecting the energy consumption of a vehicle, and the driving behavior report describes the pros and cons of the energy-saving score.

Images