TWI687827B - System and method for automatically collecting driving information - Google Patents

Abstract

This invention provides a system and a method of automatically collecting driving information comprising a waveform sampling module for obtaining an acceleration value waveform signal of the accelerometer three-axis and then converting the acceleration value waveform signal into a signal vector magnitude (SVM) waveform; a waveform smoothing module with a filtering algorithm for filtering the high frequency noise of the signal vector magnitude waveform by using a filtering algorithm; a user state recognition module for identifying the current motion state of a user by analyzing and comparing the signal vector magnitude waveform processed by the waveform smoothing module; and a control module executing a control algorithm by using the current state of motion of a user identified by the user state recognition module to automatically start a driving information collection module to collect driving information or stop collecting. This invention also discloses a method for automatically collecting driving information.

Description

System and method for automatically collecting driving information

The invention relates to a technology for collecting driving information, in particular to a system and method for automatically collecting driving information using a mobile device.

With the advent of the Internet of Vehicles era, the most practical and viable industry chain derived from the Internet of Vehicles is the "UBI Internet-based Insurance" (UBI).

UBI Internet of Vehicles Insurance is a new type of car insurance that connects smartphones to collect driving data, analyzes the driving behavior of insured persons, and then customizes the premium based on the analysis results. The company's compensation rate is significantly different from traditional auto insurance that uses only a few fixed variables (such as driver's gender, age, vehicle type, model, vehicle age, and exhaust volume) to calculate the premium, but it does not reflect the real driving behavior.

However, how to provide an effective record of the user's real driving behavior has become an important issue for many system developers.

In order to solve at least the above problems, this case provides an automatic collection of driving expenses The signal system includes: a waveform sampling module, which is used to obtain an accelerometer three-axis acceleration value waveform signal, and then convert the acceleration value waveform signal into a signal vector intensity (SVM; Signal Vector Magnitude) waveform; waveform The smoothing module has a filtering algorithm for filtering the high-frequency noise of the signal vector intensity waveform of the waveform smoothing module by using the filtering algorithm; the user state recognition module detects the waveform The signal vector intensity waveform filtered by the smoothing module is analyzed, compared, and similarity calculated to identify a user's current motion state; and the control module utilizes the user state identification module Recognizing the current movement state of the user, a control algorithm is executed to automatically start the vehicle information collection module to collect the vehicle information or stop the collection.

The present invention further proposes a method for automatically collecting driving information, which includes: acquiring an acceleration value waveform signal of an accelerometer in three axes, and then converting the acceleration value waveform signal into a signal vector intensity waveform; and filtering the signal using a filtering algorithm The high-frequency noise of the vector intensity waveform; the filtered vector intensity waveform of the signal is analyzed and compared with the similarity calculation to identify a user's current motion state; and use the identified user's current motion The state executes a control algorithm to automatically collect a line of vehicle information or stop collecting.

In the aforementioned system and method for automatically collecting driving information, it includes a user state waveform database, which is used to store the corresponding signal vector intensity waveform generated by the accelerometer under different motion states of the user.

In the aforementioned system and method for automatically collecting driving information, repacking Including the user state learning module, which has a waveform synthesis function, is based on the user's current motion state recognized by the user state recognition module, and will be stored in the user state waveform database to represent the state The corresponding waveform and the signal vector intensity waveform filtered by the waveform smoothing module are synthesized into a new signal vector intensity waveform, and then the synthesized new signal vector intensity waveform is stored in the user state waveform database.

In the aforementioned system and method for automatically collecting driving information, the user state recognition module also includes a waveform feature parameter detection unit for feature detection of the signal vector intensity waveform to obtain waveform feature parameters value.

In the aforementioned system and method for automatically collecting driving information, the user state recognition module also includes a feature similarity calculation unit for the waveform feature parameters obtained by the waveform feature parameter detection unit and the usage The waveform characteristic parameters of the signal vector intensity waveform of various motion states in the state waveform database are calculated for similarity, so that the current motion state of the user is judged by the similarity.

As can be seen from the above, the system and method for automatically collecting driving information of the present invention mainly use the built-in accelerometer (Accelerometer) of the mobile communication device (such as a smart phone, tablet computer, PDA) to automatically identify the user's current When the user is in the driving state (for example, stationary, walking, running, cycling, driving), when the user is in the driving state, the driving information is automatically collected; when in the other state, the method for automatically collecting the driving information is stopped. In addition to being able to identify the current movement state of the user, it also has a learning mechanism to achieve the purpose of personalized identification. So, without user intervention, there will be no Occurrence of missing driving information does not require additional hardware equipment, so it can save expensive costs.

In addition, the system and method for automatically collecting driving information of the present invention can be combined with the key technology of the popular Internet of Vehicles application-UBI car insurance products to reduce the cost of driving data collection, which is conducive to establishing a new business cooperation model with insurance companies.

1‧‧‧wave sampling module

2‧‧‧wave smoothing module

3‧‧‧User status recognition module

31‧‧‧ waveform characteristic parameter detection unit

32‧‧‧ Feature similarity calculation unit

4‧‧‧Control module

5‧‧‧ Driving information collection module

6‧‧‧User state waveform database

7‧‧‧User status learning module

9‧‧‧Mobile device

91‧‧‧Accelerometer

W1~W3‧‧‧Waveform

S1~S48‧‧‧Step

t‧‧‧point in time

The specific embodiments disclosed in this case will be described in detail with the following drawings, and these descriptions are shown in the following drawings:

FIG. 1 is a schematic diagram of a mobile device with a built-in accelerometer in the system for automatically collecting driving information of the present invention.

FIG. 2 is a schematic diagram of the system architecture of the present invention for automatically collecting driving information.

Figure 3 is a flow chart of the method for automatically collecting driving information of the present invention.

FIG. 4 is a flowchart of the acceleration value waveform signal of the present invention converted into a signal vector intensity waveform.

FIG. 5 is a schematic diagram of the waveform smoothing sampling method of the present invention.

Fig. 6 is a schematic diagram of waveform characteristic parameter values of the present invention.

Figures 7a to 7h are schematic diagrams of the present invention using the attribution function of fuzzy theory to calculate the similarity of waveform characteristic parameters.

FIG. 8 is a logic flow chart of the method for automatically collecting driving information of the present invention.

Figure 9 is a logic flow diagram of the method for turning off automatic collection of driving information of the present invention.

FIG. 10 is a schematic diagram of waveform synthesis of the automatic collecting driving information system of the present invention.

The following describes the implementation of the present invention by specific specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification.

It should be noted that the structure, ratio, size, etc. shown in the drawings of this specification are only used to match the content disclosed in the specification, for those who are familiar with this skill to understand and read, not to limit the implementation of the present invention The limited conditions do not have technical significance. Any modification of structure, change of proportional relationship or adjustment of size should still fall within the scope of the invention without affecting the efficacy and the purpose of the invention. The technical content disclosed by the invention can be covered. At the same time, the terms such as "front", "back" and "one" quoted in this specification are only for the convenience of description, and are not used to limit the scope of the invention, the change of the relative relationship or Adjustments, without substantial changes in the technical content, should be regarded as the scope of the invention.

Please refer to FIG. 1 and FIG. 2, which are schematic diagrams of a mobile device with an accelerometer and a system architecture diagram for automatically collecting driving information in the system for automatically collecting driving information of the present invention.

As shown in FIG. 1, the present invention first collects waveform signals generated by user activities through a built-in accelerometer 91 through a mobile device 9 carried by the user, wherein the waveform signal is the three-axis of the accelerometer 91 ( X axis, Y axis, Z axis) acceleration value waveform signal. The mobile device 9 may be an electronic device such as a tablet computer, a smart phone, or a personal digital assistant (PDA). This is not the limit.

As shown in FIG. 2, it is a schematic diagram of the system architecture of the present invention for automatically collecting driving information. The system for automatically collecting driving information includes a waveform sampling module 1, a waveform smoothing module 2, a user state recognition module 3, a control module 4, a driving information collection module 5, a user state waveform database 6 and use者国家培训模型7.

The waveform sampling module 1 is used to obtain the acceleration value waveform signal of the three axes (X, Y, Z) of the accelerometer 91, and then convert the acceleration value waveform signal into a signal vector intensity waveform. As shown in the flow chart of the conversion of the acceleration value waveform signal into the signal vector intensity waveform in FIG. 4, the waveform sampling module 1 has accelerometer detection, accelerometer software simulation, and accelerometer data sampling functions.

First, in step S11: detect the accelerometer to determine whether the mobile device has an accelerometer. In detail, it uses the accelerometer detection function to call the mobile device to manage the sensor objects to obtain the types and names of all built-in sensors, and then determine whether the accelerometer 91 exists.

If the accelerometer does not exist, the accelerometer software simulation function is activated. As shown in step S14, the software is used to simulate an accelerometer 91. In other words, the present invention can use other sensors to work together or call a software method provided by the mobile device to simulate the accelerometer.

In step S12: the acceleration signal waveform signals of the three axes (X, Y, Z) of the accelerometer 91 are sampled.

In step S13: convert the acceleration value waveform signal into a signal vector intensity waveform. Then, the signal vector intensity waveform is transmitted to the waveform smoothing module 2. Among them, the formula for converting SVM from the three axes (X, Y, Z) of the accelerometer is:

The waveform smoothing module 2 has a filtering algorithm, and uses the filtering algorithm to filter the high-frequency noise of the signal vector intensity waveform. As shown in the schematic diagram of the waveform smoothing sampling method of the present invention shown in Fig. 5, the filtering algorithm is divided into two stages: (1) sliding window size setting, and (2) signal value calculation. The purpose of "sliding window size setting" is to take out the signal vector intensity waveform and sample it at n time points before and after the time point t, and balance the numerical error of a single time point by the data change trend of a longer time (n time points). Then "signal value calculation"; the purpose of "signal value calculation" is to sample the aforementioned n time points for weight calculation,

The calculation results of the n time points are used to replace the value at the original time point t, so as to reduce the single-point value anomaly caused by high-frequency noise.

The user state recognition module 3 analyzes and compares the signal vector intensity waveform processed by the waveform smoothing module 2 and performs similarity calculation to identify a user's current motion state. The preset motion states of the present invention are five states such as stationary, walking, running, cycling, and driving. According to the detected acceleration value, the current movement state of the user is determined. For example, if the acceleration value is 0, the user is in a stationary state.

User status waveform database 6 is used to store user The intensity waveform of the corresponding signal vector generated by the accelerometer 91 in the dynamic state. In some embodiments, these signal vector intensity waveform data generation methods can be obtained from a small range of test population by capturing waveform data in various states and then averaging them to serve as a predictor of the user state waveform database 6 Set value.

In some embodiments, the user state recognition module 3 further includes a waveform feature parameter detection unit 31 and a feature similarity calculation unit 32. The waveform feature parameter detection unit 31 performs feature detection on the signal vector intensity waveform to obtain the waveform feature parameter value. As shown in the schematic diagram of the waveform characteristic parameter values in Figure 6, the waveform characteristic parameter values can be the first peak, the first peak rise time, the first peak fall time, the waveform cycle time, the interval between adjacent peaks, the second peak, 8 waveform characteristic parameter values such as second peak rise time and second peak fall time.

The characteristic similarity calculation unit 32 performs similarity calculation on the waveform characteristic parameters obtained by the waveform characteristic parameter detection unit 31 and the waveform characteristic parameters of the signal vector intensity waveforms of various motion states in the user state waveform database 6, by The similarity is used to determine the current movement state of the user. Among them, the calculation method of similarity adopts the membership function of fuzzy theory, the similarity is used as a value between 0~1, and then the similarity value of each feature is weighted to obtain the similarity representing the overall waveform. After that, the current movement state of the user (still, walking, running, cycling, driving) is estimated based on the similarity, and the movement state is transmitted to the control module 4.

Figures 7a to 7h are schematic diagrams of the present invention using the attribution function of fuzzy theory to calculate the similarity of waveform characteristic parameters.

For example, the calculation method of similarity is as follows: The waveform characteristic parameters representing the walking state in the user state waveform database are taken as examples. (Figure 7a~7h).

The eight waveform feature parameter values obtained through the waveform feature parameter detection unit 31 are respectively brought into corresponding attribution functions to obtain the similarity of each feature of the walking state waveform.

Such as: characteristic parameter values {1.85, 1.5, 10, 1.81, 4.3, 7.2, 1.9, 6.8} similarity values {0.8, 0.7, 0.85, 0.47, 0.58, 0.8, 0.89, 0.52}

Reweight these 8 feature similarity values to obtain the overall similarity to the walking waveform.

For example: the overall similarity of walking waveform = 0.8*m1+0.7*m2+0.85*m3+0.47*m4+0.58*m5+0.8*m6+0.89*m7+0.52*m8

Figures 8 and 9 are respectively logic flow charts of the method for starting and closing the automatic collection of driving information of the present invention. The control module 4 uses the current movement state of the user identified by the user state identification module 3 to execute a control algorithm to automatically start the trip information collection module 5 to collect trip information or stop the collection.

As shown in FIG. 8, the logic of the method for automatically collecting driving information of the present invention is as follows:

In step S41, the user state recognition module is presumed to be driving state.

In step S42, the GPS sensor is activated to detect the speed.

In step S43, it is confirmed whether it is a driving state based on the speed. In detail, the GPS sensor of the mobile device 9 is turned on first to detect the current speed, so as to reconfirm that the current speed does indeed conform to the driving state rather than the misjudgment of the user state recognition module 3.

In step S44, it is determined whether the driving information collection module has been activated.

In step S45, the driving information collection module is activated. If it has been activated, no action will be taken; if it has not been activated, the driving information collection module will be activated.

As shown in FIG. 9, the logic of the method for turning off automatic collection of driving information of the present invention is:

In step S46, when the user state recognition module presumes the non-driving state.

In step S47, it is determined whether the driving information collection module has been turned off.

In step S48, the driving information collection module is turned off, and the driving information is stored in the mobile device.

In some embodiments, the driving information collecting module 5 may include a GPS data receiving and calculating unit, an acceleration value calculating unit, and a driving information timing storage unit. The GPS data receiving and computing unit is responsible for recording the GPS data track of the entire driving route and calculating various driving information (for example: driving distance, driving time, direction and speed); the acceleration value calculation unit is provided according to the accelerometer 91 of the mobile device 9 The three-axis acceleration value of the car calculates sharp acceleration, sharp braking and sharp turning events; the driving information timing storage unit is to avoid If the mobile communication device loses driving information due to an unexpected situation (for example, crash or power off), all the above driving information will be stored in the memory space of the mobile device 9 regularly (for example: 30 seconds).

FIG. 10 is a schematic diagram of waveform synthesis of the system for automatically collecting driving information of the present invention. As shown in FIG. 10, the user state learning module 7 has a waveform synthesis function, and can recognize the current movement state of the user according to the user state recognition module 3 and store it in the user state waveform database 6 The corresponding waveform (such as W3) representing the state is synthesized with the signal vector intensity waveform (such as W1) filtered by the waveform smoothing module 2 to adjust the waveform type (such as W2) and the synthesized new signal vector The intensity waveform is stored in the user state waveform database 6.

Among them, the waveform synthesis method is described as follows: using Z=m1 * X+m2 * Y, where: X represents the signal vector intensity waveform W1 of the filtered high-frequency noise, and Y represents the corresponding waveform data in the user state waveform database 6 W3, Z are the synthesized new waveform W2: m1 and m2 are weight parameters, which will be adjusted according to different states (stationary, walking, running, cycling, driving).

Whenever the user status recognition module 3 completes the recognition, the waveform synthesis is performed immediately. After a period of time, the fused waveform will gradually evolve into a user-specific exclusive waveform. This personalized waveform will help improve user status recognition. The accuracy of module 3 judgment.

The invention further provides a method for automatically collecting driving information, and the method The flow chart is shown in Figure 3.

In step S1, an accelerometer 91 three-axis (X, Y, Z) acceleration value waveform signal is obtained, and then the acceleration value waveform signal is converted into a signal vector intensity waveform.

In step S2, a filtering algorithm is used to filter the high frequency noise of the signal vector intensity waveform.

In step S3, the processed signal vector intensity waveform is analyzed, compared, and similarity calculated to identify a user's current motion state.

In step S4, a control algorithm is executed, using the identified user's current movement status, to automatically collect driving information or stop collecting.

In summary, the system and method for automatically collecting driving information of the present invention mainly utilizes the built-in accelerometer (Accelerometer) and GPS sensing of mobile communication devices (such as smart phones, tablet computers, PDAs) The device automatically recognizes the current movement state of the user (for example: still, walking, running, cycling, driving), when the user is in the driving state, it automatically collects driving information; when in other states, it automatically stops collecting the driving information. method. In addition to being able to identify the current movement state of the user, it also has a learning mechanism to achieve the purpose of personalized identification. In this way, there is no need for user intervention, there will be no missing recorded driving information, and no additional hardware equipment is required, which can save expensive costs.

In addition, the system and method for automatically collecting driving information of the present invention can be combined with the key technology of the popular Internet of Vehicles application-UBI car insurance products to reduce The low-cost driving information search cost is conducive to establishing a new business cooperation model with insurance companies.

The above embodiments are used to exemplify the principles and effects of the present invention, rather than to limit the present invention. Anyone who is familiar with this skill can modify the above embodiments without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the rights of the present invention should be as listed in the scope of patent application mentioned later.

1‧‧‧wave sampling module

2‧‧‧wave smoothing module

3‧‧‧User status recognition module

31‧‧‧ waveform characteristic parameter detection unit

32‧‧‧ Feature similarity calculation unit

4‧‧‧Control module

5‧‧‧ Driving information collection module

6‧‧‧User state waveform database

7‧‧‧User status learning module

9‧‧‧Mobile device

Claims (8)

A system for automatically collecting driving information includes: a waveform sampling module for acquiring an acceleration value waveform signal of three axes of an accelerometer, and then converting the acceleration value waveform signal into a signal vector intensity (SVM) waveform; waveform The smoothing module has a filtering algorithm to filter the high-frequency noise of the signal vector intensity waveform of the waveform smoothing module using the filtering algorithm; the user state recognition module smoothes the waveform The signal vector intensity waveform filtered by the module is analyzed, compared, and similarity calculated to identify a user's current motion state; and the control module is identified by the user state identification module Of the current movement state of the user, execute a control algorithm to automatically start the traffic information collection module to collect the traffic information or stop the collection; wherein, the user state recognition module also includes a waveform feature parameter detection unit, which is It is used to perform feature detection on the signal vector intensity waveform to obtain waveform characteristic parameter values. The system for automatically collecting driving information as described in item 1 of the patent scope includes a user state waveform database, which is used to store the corresponding signal vector generated by the accelerometer of the user in different motion states Intensity waveform. The system for automatically collecting driving information as described in item 2 of the patent scope includes a user state learning module, which has a waveform synthesis function, and is based on the user identified by the user state identification module. The previous motion state, the corresponding waveform representing the state stored in the user state waveform database and the signal vector intensity waveform filtered by the waveform smoothing module are synthesized into a new signal vector intensity waveform, and then the The synthesized new signal vector intensity waveform is stored in the user state waveform database. The system for automatically collecting driving information as described in item 2 of the patent scope, wherein the user state recognition module also includes a feature similarity calculation unit for the waveform obtained by the waveform feature parameter detection unit The characteristic parameters are similar to the waveform characteristic parameters of the signal vector intensity waveform of various motion states in the user state waveform database, and the current motion state of the user is judged by the similarity. A method for automatically collecting driving information includes: acquiring an acceleration value waveform signal of three axes of an accelerometer, and then converting the acceleration value waveform signal into a signal vector intensity waveform; and filtering the signal vector intensity waveform using a filtering algorithm High-frequency noise; analyze and compare the filtered signal intensity waveform with similarity calculation to identify the current movement state of a user; and use the identified current movement state of the user to perform A control algorithm to automatically collect the vehicle information or stop the collection; wherein, the step of filtering the high-frequency noise of the signal vector intensity waveform using a filtering algorithm includes the feature detection of the signal vector intensity waveform to obtain Waveform characteristic parameter value. The method for automatically collecting driving information as described in item 5 of the patent scope includes the use of a user state waveform database to store the corresponding signal vector intensity waveform generated by the user under different motion states. The method for automatically collecting driving information as described in item 6 of the scope of the patent application includes, according to the current movement state of the identified user, the corresponding waveform and the corresponding waveform stored in the user state waveform database representing the state and The filtered signal vector intensity waveform is synthesized into a new signal vector intensity waveform, and then the synthesized new signal vector intensity waveform is stored in the user state waveform database. The method for automatically collecting driving information as described in item 6 of the scope of the patent application includes the similarity of the obtained waveform characteristic parameter with the waveform characteristic parameter of the signal vector intensity waveform of various motion states in the user state waveform database Degree calculation to determine the current movement state of the user based on the similarity.

Images