JP5187850B2 - Road surface state detection method and road surface state detection device - Google Patents

Description

  The present invention relates to a road surface state detection method and a road surface state detection device that detect a road surface state on which a vehicle travels.

  2. Description of the Related Art Conventionally, an apparatus that estimates a road surface condition (hereinafter, appropriately abbreviated as a road surface condition) on which a vehicle such as an automobile is traveling has been disclosed (for example, Patent Document 1). Specifically, the device described in Patent Document 1 estimates the road surface condition (road surface roughness) according to the internal pressure of the tire.

Japanese Patent Laying-Open No. 2005-249525 (pages 6-7, FIG. 5)

  However, the road surface roughness estimation apparatus described above includes pressure sensors 35 to 38 that detect the internal pressure P of the tire, and wheel speed sensors 31 to 34 that detect wheel speed signals of the wheels, respectively. That is, the road surface roughness estimation device is incorporated in a vehicle. Further, tire internal pressure, wheel speed, suspension resonance frequency, and the like are used for estimation of road surface conditions. Since detected values such as tire internal pressure, wheel speed, and suspension resonance frequency are expected to vary depending on the vehicle type, tire type, and suspension type, it is difficult to perform calibration before operating the road surface roughness estimation device.

  Generally, in calculating a target estimated value, the fewer the parameters related to the estimated value, the simpler the operation and the more accurate value can be obtained. Therefore, a new index that can directly know the road surface condition is also required in the estimation of the road surface condition.

  Then, an object of this invention is to provide the road surface condition detection method and road surface condition detection apparatus which can represent a road surface condition more directly.

  In order to solve the above-described problem, the first feature of the present invention has the following feature. A road surface condition detection method for detecting a road surface condition when a vehicle travels on a predetermined road surface, wherein an electric signal is transmitted from an electrode disposed at a frontal pole portion of the occupant based on the International Law 10-20. A signal acquisition step of acquiring, an analysis step of performing a time-frequency analysis on the electrical signal according to an elapsed time from the time when the travel was started, and a frequency component of the electrical signal obtained by the time-frequency analysis An optimization process for selecting a frequency component from the Nth group by applying an optimization process to the Nth group using a fitness function, and a multiple discrimination for the (N + 1) th group consisting of the frequency components selected in the optimization process An analysis step of calculating a discriminant median representing the likelihood of each frequency component included in the N + 1th group by performing an analysis, and the discriminant median calculated in the analysis step is Substituting into the fitness function, the optimization step is executed again, and the optimization step is executed again. In the optimization step, the individual included in the N + 1th group using the fitness function into which the discriminant probability is substituted. The frequency components are selected from the frequency components, and the multiple discriminant analysis is performed on the N + 2 group consisting of the frequency components selected in the optimization step that is executed again, whereby the likelihood of each frequency component included in the N + 2 group is estimated. Calculating a discriminatory probability representing a degree, and repeatedly executing the analysis step and the optimization step to determine a combination of frequency components of the electrical signal selected as an index representing the road surface condition. The gist.

  According to the first aspect of the present invention, an optimization process using an fitness function for an electrical signal acquired from an electrode arranged at the frontal pole portion of an occupant based on the International Law 10-20, and a multiple discriminant analysis Are executed repeatedly in combination. Thereby, a road surface condition can be directly expressed by using only the frequency extracted from the signal acquired from the passenger | crew.

  The second feature of the present invention relates to the first feature of the present invention, and the gist of the present invention is that the signal acquisition step acquires an electrical signal from an electrode arranged at Fp1 based on the International Law 10-20. .

  According to the second feature of the present invention, since an electric signal is acquired from one electrode arranged at Fp1, the burden on the passenger is reduced. Moreover, since an electrode should just be attached to one place, it can prevent that a measurement result becomes inaccurate under the influence of the stress of mounting | wearing a measuring apparatus.

  A third feature of the present invention relates to the first feature of the present invention, and is summarized in that the fitness function is a genetic algorithm.

  By using the genetic algorithm and the multiple discriminant analysis together, it is possible to reliably extract a component that reacts to the road surface condition from among the frequency components indicating the acquired electrical signal.

  According to a fourth aspect of the present invention, there is provided a road surface condition detecting device for detecting a road surface condition when a vehicle travels on a predetermined road surface, wherein the frontal pole portion of the occupant is based on the International Law 10-20. A signal acquisition unit that acquires an electrical signal from an electrode disposed on the electrode, an analysis unit that performs time-frequency analysis on the electrical signal according to an elapsed time from the time when the travel is started, and the time-frequency analysis. An optimization calculation unit that selects a frequency component from the Nth group by performing an optimization process using an fitness function on the Nth group of frequency components of the electrical signal, and is selected by the optimization calculation unit. An analysis unit that calculates a discriminant probability representing the likelihood of each frequency component included in the N + 1th group by performing a multiple discriminant analysis on the (N + 1) th group consisting of the frequency components. Calculated Substituting the discriminant probability into the fitness function, re-execution of the optimization process in the optimization calculation unit, and in the optimization process executed again, the fitness with the discriminant probability substituted By selecting a frequency component from the individual frequency components included in the N + 1th group using a function, and performing a multiple discriminant analysis on the N + 2th group consisting of the frequency components selected in the optimization process executed again. It is selected by calculating a discriminant probability representing the likelihood of each frequency component included in the (N + 2) th group, and repeatedly executing a multiple discriminant analysis in the analysis unit and an optimization process in the optimization calculation unit. The gist of the invention is to determine a combination of frequency components of the electrical signal as an index representing the road surface condition.

  ADVANTAGE OF THE INVENTION According to this invention, the road surface condition detection method and road surface condition detection apparatus which can express a road surface condition more directly can be provided.

FIG. 1 is a schematic diagram for explaining a road performance tire performance evaluation test performed using a road surface condition detection device. FIG. 2 is a configuration diagram illustrating the road surface condition detection device. FIG. 3 is a flowchart for explaining a road surface condition detection method. It is a schematic diagram explaining an example of the test course which makes a vehicle drive | work. FIG. 4 is a flowchart for specifically explaining step S <b> 2 in FIG. 3. FIG. FIG. 6 is a diagram illustrating an example of a gene expression (first generation) initially assigned to a plurality of frequency bands (4 Hz to 22 Hz). FIG. 7 is a diagram illustrating an example of a frequency band that characterizes road surface conditions.

  Next, an embodiment of a road surface condition detection method and a road surface condition detection device according to the present invention will be described with reference to the drawings.

  Specifically, (1) the configuration of the road surface condition detection device, (2) the description of the road surface condition detection method, (3) the action and effect, (4) other embodiments will be described.

(1) Configuration of Road Surface Condition Detection Device The road surface condition detection device 1 will be described with reference to FIG. Specifically, (1-1) a test for acquiring an electric signal from an occupant and (1-2) a configuration of a road surface condition detection device will be described.

(1-1) Test for Acquiring an Electric Signal from an Occupant FIG. 1 is a schematic diagram for explaining a road performance tire performance evaluation test performed using the road surface condition detection device 1.

  The road surface condition detection apparatus 1 includes an evaluation apparatus body 10 and a signal acquisition unit 20. The road surface condition detection device 1 represents a road surface condition using an electrical signal that can be acquired from the occupant 3 when the vehicle 100 on which the tire 101 is mounted travels on a predetermined road surface (described later). Details of the evaluation apparatus body 10 will be described later.

  The signal acquisition part 20 acquires an electrical signal from the electrode arrange | positioned at the frontal pole part of the passenger | crew 3 based on the international law 10-20 method. The signal acquisition unit 20 includes an electrode 21 and an electrode 22. The electrode 21 is arrange | positioned at Fp1 and acquires the electric signal in Fp1. The electrode 22 is a reference electrode. The electrode 22 is arrange | positioned at an earlobe etc., for example. The difference between the potential at the electrode 21 and the potential at the electrode 22 is amplified by a differential amplifier. Thereby, noise can be removed and a weak electroencephalogram signal can be accurately extracted.

  The occupant 3 gets on the vehicle 100 with the signal acquisition unit 20 mounted at the Fp1 location. The road surface condition detection apparatus 1 acquires an electric signal that can be acquired from the Fp1 of the occupant 3 via the signal acquisition unit 20 while the vehicle travels on a predetermined test course.

(1-2) Configuration of Road Surface Condition Detection Device FIG. 2 is a configuration diagram illustrating the road surface condition detection device 1. The road surface condition detection apparatus 1 includes an evaluation apparatus body 10 and a signal acquisition unit 20. The signal acquisition unit 20 is an electrode. The signal acquisition unit 20 is attached to Fp1 of the frontal pole portion of the occupant 3.

  The evaluation apparatus body 10 includes a signal input unit 11 to which the electric signal acquired by the signal acquisition unit 20 is supplied, a calculation unit 12 that calculates a road surface condition from the electric signal supplied to the signal input unit 11, and a calculation unit 12 And a display unit 13 for displaying the result of the analysis.

  The evaluation apparatus main body 10 also includes an input unit 14 such as a keyboard and a mouse that receives input from the user, and a storage unit 15 including a hard disk drive, a semiconductor memory, and the like.

  Specifically, the calculation unit 12 includes a frequency analysis calculation unit 121, an optimization calculation unit 122, and an analysis calculation unit 123.

  The frequency analysis calculation unit 121 performs time frequency analysis of the electrical signal. In psychology, electroencephalograms are roughly classified into four types, δ waves, θ waves, α waves, and β waves, depending on the frequency band. For example, the δ wave is an electroencephalogram having a frequency band of less than 4 Hz. The δ wave does not appear when awake, but frequently appears during deep sleep. The θ wave is an electroencephalogram having a frequency band of 4 Hz or more and less than 8 Hz. Appears frequently during light sleep. The α wave is an electroencephalogram having a frequency band of 8 Hz or more and less than 13 Hz. Appears conspicuously on the occipital side when the eyes are resting, and the appearance is suppressed when the eyes are opened. It is also an index for evaluating relaxation. The β wave is an electroencephalogram having a frequency band of 13 Hz or higher. It is seen predominantly in the frontal and temporal regions. There is little regularity.

  In the present embodiment, the electroencephalogram in the frequency band described above is separated by analyzing the electrical signal acquired from Fp1. As the frequency analysis, fast Fourier transform, wavelet transform, or the like can be used. In this embodiment, fast Fourier transform is used.

  Since the electroencephalogram is a time series signal in which a plurality of factors are intertwined in a complicated manner, it is difficult to extract the feature quantity of the electroencephalogram only by frequency analysis. In this embodiment, features are extracted using multivariable analysis.

  The optimization calculation unit 122 selects a frequency component from the Nth group by performing an optimization process using the fitness function on the Nth group including the frequency components of the electrical signal obtained by the time-frequency analysis. In the present embodiment, the optimization calculation unit 122 uses a genetic algorithm as an example of an optimization processing technique. Optimization processing is performed on the frequency component of the electrical signal obtained by the time-frequency analysis.

  The analysis calculation unit 123 performs a discriminant analysis on the (N + 1) th group including the frequency components selected by the optimization calculation unit 122, thereby obtaining a discriminant median representing the likelihood of each frequency component included in the (N + 1) th group. calculate.

  In the road surface condition detection apparatus 1, the discriminant probability calculated in the analysis calculation unit 123 is substituted into the fitness function, and the optimization calculation unit 122 executes the optimization process again. In the optimization process executed again, the frequency component is selected from the individual frequency components included in the (N + 1) th group using the fitness function into which the discriminant probability is substituted.

  In the optimization process executed again, a multiple discriminant analysis is performed on the (N + 2) th group composed of frequency components selected from the (N + 1) th group. As a result, a discriminant median representing the likelihood of each frequency component included in the (N + 2) th group is calculated.

  In the road surface condition detection apparatus 1, the combination of the frequency components of the finally extracted electric signal is obtained by repeatedly executing the multiple discriminant analysis in the analysis calculation unit 123 and the optimization process in the optimization calculation unit 122. Determined as an indicator of the situation.

(2) Description of Road Surface Condition Detection Method Next, the road surface condition detection method will be described with reference to the drawings. Specifically, (2-1) description of the entire road surface condition detection method, (2-2) step S1, (2-3) step S2, and (2-4) step S3 will be described.

(2-1) General Description of Road Surface Condition Detection Method FIG. 3 is a flowchart illustrating the road surface condition detection method. As shown in FIG. 3, the road surface condition detection method includes steps S1 to S3. Step S1 is a signal acquisition process. In step S <b> 1, an electrical signal is acquired from an electrode disposed on the frontal pole portion of the occupant 3. Step S2 is an analysis process. In the analysis step, the fitness function is applied to a group consisting of an analysis step for performing time frequency analysis on the electric signal according to an elapsed time from the time when the running is started, and a frequency component of the electric signal obtained by the time frequency analysis. Represents the likelihood of each frequency component by performing a multiple discriminant analysis on the optimization step of selecting frequency components by performing optimization processing using and the group of frequency components selected in the optimization step And an analysis step for calculating the discriminant predictive value. Step S3 is a determination step for determining an indicator of road surface conditions. Details of the processing in step S2 will be described later.

(2-2) Step S1: Process for Acquiring an Electric Signal from an Occupant In step S1, a process for acquiring an electric signal from an occupant will be specifically described. The occupant travels on the test course in a vehicle in which the evaluation target tire is mounted with the electrode mounted on the frontal pole portion.

  The test course has a general paved road, a rough road surface that is not flatter than a general paved road, a wavy road surface, a narrow road, cornering, a wet road surface, a snowy road surface, a frozen road surface, and the like. For general paved roads, the driving stability such as straight running stability, lane change performance and cornering stability will be evaluated. For rough roads and wavy roads, steering stability on rough roads is evaluated. In Kushiro, we will evaluate the handling stability in the Kushiro such as wobbling. For wet roads, snowy roads, and frozen roads, steering stability on wet and icy roads will be evaluated. In addition, noise and vibration on general paved roads and rough roads, riding comfort on steps, and riding comfort on uneven roads will be evaluated. Furthermore, the braking performance and acceleration / deceleration performance are evaluated.

  FIG. 4 is a schematic diagram illustrating an example of a test course. The test course 200 includes at least a straight paved straight road surface R1, a rough road surface R2, a rough road surface R3 having a undulation larger than R2, a narrow road R4, a corner R5 having a large bank angle, and a wavy road surface. R6 and a corner R7 having a small bank angle.

  When the vehicle 100 reaches each section (R1 to R7) during traveling, a time stamp indicating the start of brain wave measurement may be recorded together with the brain wave. When the section has been passed, a type stamp indicating the end may be recorded together with the electroencephalogram. This makes it easier to extract brain waves during a period in which the vehicle travels in a predetermined section (a period from the start time stamp to the end time stamp).

(2-3) Step S2: Analysis Processing Next, specific steps executed in step S2 will be described with reference to FIG. Step S2 further includes steps S21 to S25.

  First, in step S21, the time frequency analysis of the electric signal supplied from the signal acquisition unit 20 is performed. Specifically, an electrical signal of an electroencephalogram acquired from Fp1 after a certain period of time has elapsed from the time when the running is started is decomposed into a plurality of frequency bands (4 Hz to 22 Hz) using fast Fourier transform.

  Subsequently, in step S22, an initial gene population is determined for a plurality of frequency bands (4 Hz to 22 Hz). That is, a value of 0 or 1 generated at random is assigned to a plurality of frequency bands (4 Hz to 22 Hz). Here, it may not be 0 or 1. It may be an analog value.

  FIG. 6 shows an example of gene expression. The gene expression initially assigned to a plurality of frequency bands (4 Hz to 22 Hz) is referred to as the first generation (corresponding to the Nth group).

  In step S23, the multiple discriminant analysis is performed on the plurality of frequency bands (4 Hz to 22 Hz) having the gene expression determined in step S22, and the discriminant intermediate rate included in the plurality of frequency bands (4 Hz to 22 Hz) is calculated. To do. As the multiple discriminant analysis, discriminant analysis based on Mahalanobis general distance is used. The discriminative predictive value is substituted into the fitness function of the genetic algorithm, and the fitness of the individual genetic algorithm (first generation is the first generation) is calculated.

  In step S24, it is determined whether or not the degree of application has reached an end condition. If the end condition is reached, the process proceeds to step S3.

  On the other hand, if the applicability does not reach the end condition in step S24, the process returns to step S23, and the second generation (N + 1th group) selected from the first generation by performing the multiple discriminant analysis on the first generation. The discriminative probability of the frequency component is calculated. The discriminative predictive value is substituted into the fitness function of the genetic algorithm, and the fitness of a plurality of frequency bands (4 Hz to 22 Hz) having the second generation gene expression is calculated.

  In step S24, when the applicability does not reach the end condition, the fitness is adjusted by performing “selection” in the genetic algorithm before returning to the process of step S23. The selection includes a roulette method, a scaling method, a rank method, a tournament method, an elite preservation strategy method, and the like. Also, the gene expression is rearranged by performing “crossover” in the genetic algorithm. Crossover includes one-point crossover, multiple-point crossover, uniform crossover, and the like. In addition, gene expression may be rearranged by performing “mutation” in a genetic algorithm. Steps S23 to S25 are repeated until the end condition is reached.

(2-4) Step S3
In the determination step S3, based on the processing in the analysis step S2, a group of frequency bands represented by the finally extracted gene expression is determined as an index representing the road surface condition. FIG. 7 is a diagram illustrating an example of a frequency band that characterizes road surface conditions. In FIG. 7, the frequency band (ABDFIJKPQR) to which 1 is assigned can be treated as an index representing the road surface Rn in a certain state.

(3) Action / Effect According to the road surface condition detection apparatus 1, an optimization process using an fitness function for an electrical signal acquired from an electrode arranged at the frontal pole portion of an occupant based on the International Law 10-20 , Repeatedly executed in combination with multiple discriminant analysis. Thereby, a road surface condition can be directly expressed by using only the frequency extracted from the signal acquired from the passenger | crew.

  According to the road surface condition detection apparatus 1, an electrical signal is acquired from an electrode arranged at Fp1 based on the International Law 10-20. Thus, since an electric signal is acquired from one electrode arranged in Fp1, according to the road surface condition detection apparatus 1, the burden on a passenger | crew is reduced. Moreover, since an electrode should just be attached to one place, it can prevent that a measurement result becomes inaccurate under the influence of the stress of mounting | wearing a measuring apparatus.

(4) Other Embodiments As described above, the contents of the present invention have been disclosed through the embodiments of the present invention. However, it is understood that the descriptions and drawings constituting a part of this disclosure limit the present invention. Should not. From this disclosure, various alternative embodiments, examples and operational techniques will be apparent to those skilled in the art.

  It goes without saying that the present invention includes various embodiments not described herein. Therefore, the technical scope of the present invention is defined only by the invention specifying matters according to the scope of claims reasonable from the above description.

  DESCRIPTION OF SYMBOLS 1 ... Tire performance evaluation apparatus, 3 ... Crew, 10 ... Evaluation apparatus main body, 11 ... Signal input part, 12 ... Calculation part, 13 ... Display part, 14 ... Input part, 15 ... Memory | storage part, 20 ... Signal acquisition part, 21 DESCRIPTION OF SYMBOLS ... Electrode, 22 ... Electrode, 100 ... Vehicle, 101 ... Tire, 121 ... Frequency analysis calculation part, 122 ... Optimization calculation part, 123 ... Analysis calculation part, 200 ... Test course

Claims (5)

A road surface condition detection method for detecting the condition of the road surface when the vehicle travels on a predetermined road surface,
A signal acquisition step of acquiring an electrical signal from an electrode disposed on a frontal pole portion of an occupant of the vehicle based on international law 10-20;
An analysis step of performing a time-frequency analysis on the electrical signal according to an elapsed time from the time when the traveling was started;
An optimization step of selecting a frequency component from the Nth group by applying an optimization process to the Nth group of frequency components of the electrical signal obtained by the time-frequency analysis using an fitness function;
An analysis step of calculating a discriminant median representing the likelihood of each frequency component included in the N + 1 group by performing a multiple discriminant analysis on the N + 1 group of frequency components selected in the optimization step. Have
Substituting the discriminant probability calculated in the analysis step into the fitness function, and executing the optimization step again,
In the optimization step executed again, a frequency component is selected from the individual frequency components included in the N + 1th group using a fitness function into which the discriminant probability is substituted,
A discriminant median representing the likelihood of each frequency component included in the N + 2 group is calculated by performing multiple discriminant analysis on the N + 2 group consisting of the frequency components selected in the optimization step executed again. ,
A road surface condition detection method for determining a combination of frequency components of the electrical signal selected by repeatedly executing the analysis process and the optimization process as an index representing the road surface condition.   The road surface condition detection method according to claim 1, wherein in the signal acquisition step, an electric signal is acquired from an electrode arranged at Fp1 based on the International Law 10-20.   The road surface condition detection method according to claim 1, wherein the fitness function is a genetic algorithm. When the vehicle travels on a predetermined road surface, the road surface state detection device detects the road surface state,
A signal acquisition unit for acquiring an electrical signal from an electrode disposed on a frontal pole portion of an occupant of the vehicle based on International Law 10-20;
An analysis unit for performing a time-frequency analysis on the electrical signal according to an elapsed time from the time when the traveling was started;
An optimization calculation unit that selects a frequency component from the N-th group by performing an optimization process using an fitness function on the N-th group including the frequency components of the electrical signal obtained by the time-frequency analysis;
An analysis unit for calculating a discriminant intermediate ratio representing the likelihood of each frequency component included in the N + 1 group by performing a multiple discriminant analysis on the N + 1 group of frequency components selected by the optimization calculation unit; Have
Substituting the discriminant probability calculated in the analysis unit into the fitness function, and executing the optimization process again in the optimization calculation unit,
In the optimization process executed again, the frequency component is selected from the individual frequency components included in the N + 1th group using the fitness function in which the discriminant probability is substituted,
A discriminant median representing the likelihood of each frequency component included in the N + 2 group is calculated by performing a multiple discriminant analysis on the N + 2 group consisting of the frequency components selected in the optimization process executed again. ,
Road surface condition detection that determines a combination of frequency components of the electrical signal selected by repeatedly executing the multiple discriminant analysis in the analysis unit and the optimization process in the optimization calculation unit as an index representing the road surface condition apparatus.   The road surface condition detection device according to claim 4, wherein the signal acquisition unit is arranged at Fp1 based on the International Law 10-20.

Images