JP2009022370A - Human condition estimation system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately measure the condition of a subject to a stimulus. <P>SOLUTION: A condition estimation system 1 has a vibration generator 2 applying an awakening stimulus to a driver, an air conditioner 3, a speaker 6 and a display 7, a skin electrical activity sensor 4 acquiring biological information showing a biological reaction (SPA) of the driver to the applied awakening stimulus, and an ECU 8 setting reaction time to measure a biological reaction (measurement time) in accordance with the kind of the awakening stimulus, measuring the biological reaction on the basis of the biological information acquired during the set measurement time, and estimating the condition of the subject on the basis of the measured biological reaction. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a human state estimation device that estimates the state of a subject.

  2. Description of the Related Art Conventionally, a technique for reducing or eliminating driver drowsiness by applying a plurality of types of stimuli is known. For example, Patent Document 1 below discloses a drowsiness prevention device that calculates the number of blinks of a driver per unit time and determines the type of stimulus given to the driver based on the calculation result.

A method for estimating the effect of the device described in Patent Document 1 below, that is, the awakening effect, is also known. For example, Patent Document 2 below captures a human state change by applying a physical stimulus periodically or before an event occurs, and quantifying a biological response to the application using various physiological indices (biological information). A human state detection device is disclosed. Patent Document 3 below discloses a technique for simulating human sleepiness or wakefulness by constructing a sleepiness model using the principle of quantification.
JP-A-6-328968 Japanese Patent Application No. 2005-316773 Japanese Patent Application No. 2006-259204

  However, in the apparatus of Patent Document 3 described above, there is a possibility that the accuracy of estimation of the arousal effect is lowered depending on the type of stimulus. In general, the body's acceptance rate for painful stimuli typified by electrical stimulation is fast, and the response to involuntary physiological indices appears quickly, but pressure stimuli such as cold air and massage take time. Because it is received by somatic sensation, it is difficult to produce a fast response like electrical stimulation. In order to calculate an SPR (Skin Potential Response) value indicating a skin potential response, the apparatus described in Patent Document 3 extracts a maximum value and a minimum value of an SPR component in a predetermined section, and calculates the two values. Although the difference is calculated, in this method, the awakening effect may not be accurately grasped depending on the characteristic of the stimulus. For example, even if the device described in Patent Document 3 is used, it is difficult to treat the awakening effect equally between a stimulus having an immediate effect (immediate effect stimulus) and a stimulus having a delayed effect (delayed effect stimulus). Therefore, there is a possibility that the accuracy of estimating the awakening effect is lowered.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a human state estimation device that can accurately estimate the state of a subject with respect to a stimulus.

  The human state estimation device according to the present invention is a human state estimation device that estimates the state of a subject, a stimulus applying unit that applies a stimulus to the subject, and a biological reaction of the subject to the stimulus applied by the stimulus applying unit. Set by biometric information acquisition means for acquiring biometric information to be shown, measurement time setting means for setting a measurement time for measuring a biological reaction according to the type of stimulus applied by the stimulus applying means, and measurement time setting means A biological reaction measuring means for measuring a biological reaction based on the biological information acquired by the biological information acquiring means during the measured time, and estimating the state of the subject based on the biological reaction measured by the biological reaction measuring means And a state estimating means.

  According to such a human state estimation device, a measurement time for measuring a biological reaction is set, and based on the biological information acquired during the measurement time, the subject's response to the stimulus applied to the subject Biological reactions are measured. Subsequently, the state of the subject is estimated based on the measured biological reaction. The appearance of a biological response to a stimulus differs for each stimulus to be applied. For example, there is a biological reaction that responds quickly to a stimulus and a biological reaction that responds slowly. For this reason, setting the measurement time of the biological reaction according to the reaction time of the biological reaction is effective for accurately measuring the biological reaction. Therefore, by setting the biological reaction measurement time for each stimulus, the biological reaction for each stimulus can be measured with high accuracy, and the state of the subject with respect to the stimulus can be estimated with high accuracy. Note that the state of the subject to be estimated is, for example, an arousal state, a psychological state, a fatigue state, concentration, attention, or stress.

  In the human state estimation apparatus of the present invention, it is preferable that the types of stimuli include a first stimulus and a second stimulus having a slower biological response to the stimulus than the first stimulus.

  In this case, the measurement time is set for each of the first stimulus and the second stimulus that has a slower biological response to the stimulus than the second stimulus. For example, an immediate stimulus with a fast biological response to a stimulus (fast somatosensory reception rate and immediate effect) and a slower biological response to this stimulus (slow somatosensory acceptance rate) The measurement time is set for each of the slow-acting stimuli. Therefore, it is possible to accurately measure biological responses to these two types of stimuli, and thus to accurately estimate the state of the subject with respect to each stimulus.

  In the human state estimation apparatus according to the present invention, it is preferable that the biological reaction measuring unit measures the biological reaction by integrating the biological information acquired during the measurement time set by the measurement time setting unit.

  In this case, the biological reaction is measured by integrating the biological information acquired during the measurement time. As a result, it is possible to measure a biological reaction that reflects the reaction time of a different biological reaction for each stimulus, so the biological reaction of the subject to be converted with time transition can be measured with higher accuracy, and the state of the subject with respect to the stimulus can be more accurately measured Can be estimated.

  In the human state estimation device of the present invention, based on the biological reaction measured by the biological reaction measuring means, first index value acquisition means for acquiring a first index value corresponding to the state change amount of the subject, and the state of the subject Gain setting means for setting a gain that determines the degree of addition of the first index value when estimating the state, and the state estimation means is acquired by the first index value acquisition means according to the gain set by the gain setting means It is preferable that the degree of addition of the first index value is defined, and the state of the subject is estimated based on the first index value of the degree of addition according to the gain and the state of the subject in the past.

  In this case, the first index value is acquired based on the measured biological reaction, a gain for the first index value is set, and the degree of consideration of the first index value is defined according to the gain. Then, the state of the subject is estimated based on the first index value and the past state of the subject. The first index value is a value corresponding to various changes in the state of the subject obtained from the subject's physiological index (biological information), behavioral index, and the like, and serves as an index for estimating the state of the subject. is there. The gain is an index for determining how much the first index value is reflected in the state estimation of the subject. If the first index value is added to the state estimation of the subject as it is, the estimation accuracy decreases. Therefore, by providing a gain for the first index value and adjusting the degree of reflection of the first index value in the state estimation, the influence of noise on the state estimation is suppressed, and the state of the subject is estimated with high accuracy. can do.

  In the human state estimation device according to the present invention, the second index value acquisition that acquires a second index value that is larger in change with respect to the state change of the subject than the first index value is based on the biological reaction measured by the biological reaction measuring means. It is preferable that a gain setting unit further sets the gain based on the second index value acquired by the second index value acquisition unit.

  In this case, based on the measured biological reaction, a second index value having a larger change with respect to the change in the state of the subject than the first index value is acquired. Then, a gain is set based on the second index value. In this way, by setting the gain with an index value that responds quickly to a change in the state of the subject (an index value with a large change), the state of the subject can be quickly reflected in the gain (and thus the subject's state). The state can be quickly reflected in the state estimation), and the state estimation accuracy can be further improved.

  According to such a human state estimation device, the biological reaction of the subject is measured based on the biological information acquired at the measurement time set according to the type of stimulus, and the state of the subject is determined based on the biological reaction. Since it is estimated, the state of the subject with respect to the stimulus can be estimated with high accuracy.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same or equivalent elements are denoted by the same reference numerals, and redundant description is omitted.

  In the present embodiment, the human state estimation device according to the present invention is applied to a state estimation system 1 that is mounted on a vehicle and estimates a driver's drowsiness level (wake state).

  The state estimation system 1 gives a reference stimulus to the driver and detects an electrodermal activity (EDA), particularly a skin potential response (SPR) as a driver's biological response to the reference stimulus. Then, the SPR value is calculated based on the SPR in the predetermined reaction time among the detected SPRs. Subsequently, the state estimation system 1 calculates an AR value (first index value) based on the SPR value by an AR (Active Reference) method, and estimates a drowsiness level based on the AR value. In the AR method, a reference stimulus (vibration, etc.) is given to a person at regular intervals, and a person's state (wakefulness) is determined based on a relative change for each section of a person's biological reaction obtained in response to the stimulus. State).

  Subsequently, the state estimation system 1 gives a wake-up stimulus to alert the driver according to the drowsiness level when the drowsiness level is at a level that causes any difficulty in driving. Subsequently, the state estimation system 1 estimates the current sleepiness level based on the AR value obtained this time and the sleepiness level estimated last time by using the sleepiness estimation model. In particular, the state estimation system 1 sets a reaction time (measurement time) for measuring the SPR according to the type of stimulus (reference stimulus and wakefulness stimulus) to be applied in order to accurately measure the biological reaction. .

  With reference to FIGS. 1-3, the state estimation system 1 which concerns on this Embodiment is demonstrated. FIG. 1 is a diagram showing a configuration of a state estimation system according to the present embodiment. FIG. 2 is a graph showing an example of detected SPR components. FIG. 3 is a graph showing two types of SPR components.

  As shown in FIG. 1, the state estimation system 1 includes a vibration generator 2, an air conditioner 3, an electrical skin activity sensor 4, an amplifier 5, a speaker 6, a display 7, and an ECU (Electronic Control Unit) 8. In the present embodiment, the vibration generator 2, the air conditioner 3, the speaker 6 and the display 7 correspond to the stimulus applying means described in the claims, and the electrodermal activity sensor 4 is described in the claims. The measurement time setting means, the biological reaction measurement means, the first index value acquisition means, the gain setting means, the second index value acquisition means, and the state described in the appended claims This corresponds to the estimation means.

  The vibration generator 2 is a device that generates vibration based on a vibration generation signal from the ECU 8, and applies a fine vibration for reference stimulation and a vibration for arousal stimulation to the driver. The vibration generator 2 is built in several places on the seat of the driver's seat. The position and the number of the built-in parts are arbitrary, and for example, a total of 6 pieces are built in one place on the left and right sides of the driver's back, waist, and thigh. The vibration generator 2 can change parameters such as the intensity of vibration to be generated, frequency, and generation time. The vibration generator 2 generates vibration according to the parameter indicated by the vibration generation signal input from the ECU 8.

  In order to distinguish all vibrations generated by the vibration generating device 2 from vibrations generated in the vehicle, parameters of strength and frequency that are clearly different from those generated in the vehicle are set. In the slight vibration for the reference stimulus, a parameter having a small strength is set so as not to make the driver feel uncomfortable. Since the vibration for arousal stimulation needs to improve the driver's arousal state, a parameter with a large strength is set.

  The air conditioner 3 is a device that cools and heats the passenger compartment. The air conditioner 3 can receive an input from the driver and perform control according to the input, and parameters (for example, set temperature, air volume, wind direction, etc.) indicated by the air conditioning control signal input from the ECU 8. It is possible to give an awakening stimulus to the driver by operating based on a parameter indicating the For example, the air conditioner 3 sends a predetermined amount of cold air toward the driver's seat based on the air conditioning control signal.

  The electrodermal activity sensor 4 is a sensor that detects EDA, particularly skin potential activity (SPA). The electrodermal activity sensor 4 is attached to the left and right sides of the steering wheel with which the palm of the driver contacts in order to detect the driver's mental sweating. When the electrodermal activity sensor 4 detects SPA, the detection signal is transmitted to the amplifier 5. This detection signal indicates biological information indicating a driver's biological reaction. It is to be noted that humans tend to be prone to mental sweating with respect to stimuli when the arousal state is high, and mental sweating tends to be difficult to occur when the arousal state is low.

  The amplifier 5 is a device that amplifies the detection signal from the electrodermal activity sensor 4 and transmits the amplified detection signal to the ECU 8.

  The speaker 6 and the display 7 are used in common with each system in the vehicle. In the state estimation system 1, the speaker 6 and the display 7 are used to give an arousal stimulus to the driver. When the speaker 6 receives a sound signal from the ECU 8, the speaker 6 outputs sound according to the sound signal. When the display 7 receives the image signal from the ECU 8, the display 7 displays an image according to the image signal.

  The ECU 8 includes a central processing unit (CPU), a read only memory (ROM), a random access memory (RAM), and the like, and performs overall control of the state estimation system 1. The ECU 8 periodically transmits a vibration generation signal for generating a fine vibration for reference stimulation. Subsequently, the ECU 8 estimates the drowsiness level of the driver using the detection signal input from the amplifier 5 and a drowsiness estimation model stored in advance in a ROM or the like. Furthermore, when the sleepiness level is a level that hinders driving, the ECU 8 uses the vibration generator 2, the air conditioner 3, the speaker 6, and the display 7 to give an arousal stimulus to the driver.

  Here, a state estimation method using the sleepiness estimation model will be described. When starting the state estimation, the ECU 8 gives a reference stimulus to the driver at a constant cycle. The ECU 8 detects EDA (especially SPA) as biological information for the reference stimulus every time the reference stimulus is applied. Subsequently, the ECU 8 removes a skin potential level (SPL) that is a low frequency component from the detected SPA. The component obtained by removing the low frequency component (SPL) from the SPA is the component (biological information) of the SPR.

  For example, as shown by the solid line E in FIG. 2, the SPR component changes periodically with the passage of time. In addition, the vertical axis | shaft of the graph shown in FIG. 2 shows the value of SPR component, and a horizontal axis shows time. In this graph, the waiting time (latency) from when a stimulus (reference stimulus or arousal stimulus) is given to the driver until the start of SPR component extraction is indicated by Ta, and is detected at the latency Ta. In particular, the SPR component is indicated by a broken line Ea. The latency Ta is set based on experimental results and the like, and is stored in the ECU 8. The latency Ta stored in the ECU 8 may be one type or a plurality of types.

  Of the SPR component indicated by the solid line E, one cycle after the latency Ta has elapsed is defined as an SPR component indicating a biological response to a stimulus (reference stimulus or arousal stimulus), and the time corresponding to the one cycle is defined as a reaction time T. The reaction time T is set for each type of stimulus based on an experiment for deriving physiological response characteristics to the stimulus, and is stored in the ECU 8.

  The reaction time T will be described with an example. In general, the body's acceptance rate for painful stimuli typified by electrical stimulation is fast, and the response to involuntary physiological indices appears quickly, but pressure stimuli such as cold air and massage take time. Because it is received by somatic sensation, it is difficult to produce a fast response like electrical stimulation. Therefore, a short reaction time T is set for a stimulus in which a biological reaction appears quickly (such as a painful stimulus) (immediate effect stimulus), and a stimulus (delayed stimulus) in which a biological reaction appears over time, such as a pressure stimulus, A longer reaction time T is set than in the case of immediate-effect stimulation. That is, the reaction time T is set according to the response speed of the living body to the stimulus.

  For example, as shown in FIG. 3, the ECU 8 sets the reaction time Tp set for the stimulus applied by the vibration generator 2 and the reaction time Tq set for the stimulus applied by the air conditioner 3. And remember. The vibration applied from the vibration generator 2 is a kind of immediate effect stimulus, and the cold air given by the air conditioner 3 is a kind of slow effect stimulus. Therefore, the reaction time Tq is set longer than the reaction time Tq. Of course, the ECU 8 may store the reaction time set for the stimulus applied by the speaker 6 or the display 7.

The ECU 8 integrates the SPR component in the reaction time using the following formula (1) to derive (measure) the SPR value in the reaction time. In the equation (1), t _a is the time latency Ta from the time that impart stimulation has elapsed, t _b is the time when the reaction time has elapsed from the time t _a.

For example, when calculating the SPR value for the stimulus applied by the vibration generator 2, the ECU 8 integrates the SPR component Ep in the reaction time Tp and calculates the SPR value for the stimulus applied by the air conditioner 3. Is integrated with the SPR component Eq in the reaction time Tq. That is, the reaction times Tp and Tq are measurement times for measuring a biological reaction set according to the type of stimulus given to the driver, and t _a and t _b in the equation (1) are respectively measured. It can be said that it is a start time and a measurement end time. Moreover, the SPR component Ep in the reaction time Tp and the SPR component Eq in the reaction time Tq are biological information acquired during the measurement time (reaction time Tp, Tq).

Each time the ECU 8 calculates (measures) the SPR value, it calculates the AR value of the SPR by the following equation (2) using the SPR (t) calculated this time and the SPR (t-1) calculated last time. The AR value is a normalized value of the relative change of the SPR value, and 0 is a reference value. The AR value correlates with the drowsiness level, and it can be estimated that the drowsiness is weaker (the arousal state is higher) as it is larger than 0, and the drowsiness is stronger (the arousal state is lower) as it is smaller than 0.

  In Equation (2), AR (t) is the AR value for the current reference stimulus, SPR (t) is the SPR value for the current reference stimulus, and SPR (t−1) is the SPR value for the previous reference stimulus. It is. The time interval between t-1 and t is a reference stimulus application interval.

The ECU 8 uses a sleepiness estimation model for estimating the sleepiness level based on the AR value. This sleepiness estimation model is represented by the following formula (3).

  In Equation (3), DL (t) is the current sleepiness level, and DL (t−1) is the previous sleepiness level. In the present embodiment, the drowsiness level (wakefulness level) is expressed in stages of 6 to 1 (D0 to D5). Sleepiness level 6 (wake level D0) indicates a state in which sleepiness is the weakest (highest alertness level), and sleepiness level 1 (wake level D5) indicates a state in which sleepiness is the strongest (lowest alertness level).

  In Equation (3), K is a gain for determining the degree of reflection of the AR value for the current reference stimulus. The gain K is a value between 0 and 1, and is set in three stages based on the SPR frequency (second index value) indicating the degree of fluctuation of the SPR component. It should be noted that the number of steps for setting the gain K is not limited.

  The SPR frequency (second index value) used to set the gain K is the response when the SPR component acquired at the reaction time T set according to the type of stimulus and the previous reference stimulus are applied. It is determined based on the average value mean and standard deviation value σ of the SPR component at time T. Specifically, the sum of the number of times the SPR component has become larger than the value mean + 3σ and the number of times the SPR component has become smaller than mean−3σ in the reaction time T is defined as the SPR frequency. Since the SPR frequency directly represents the degree of fluctuation of the SPR component, the change with respect to the driver's state change is larger than the AR value.

  When the SPR frequency is high, the driver is in a sympathetic nerve activated state. In other words, the driver is said to be in a state of low sleepiness (high arousal level) and high activity level from the aspect of behavioral state, and from the aspect of psychological state that the tension state is high and excited state. I can say that. In such a case, since the response to the stimulus is sensitive, there is a case where the AR value is greatly fluctuated due to a high sensitivity to a slight noise (high possibility of picking up noise). In this case, since the influence of noise on the drowsiness level DL is increased, the gain K is decreased and the degree of reflection of the AR value is decreased. On the other hand, when the SPR frequency is low, it can be said that the driver is sleepy in terms of behavioral state (low wakefulness) and low in activity level, and in terms of psychological state, the driver has low tension. It can be said that it is in a relaxed state. In such a case, since the response to the stimulus is insensitive, the response to noise is also slow. Therefore, the gain K is increased and the reflection degree of the AR value is increased.

  Specifically, threshold values th1 and th2 for the SPR frequency are prepared. The thresholds th1 and th2 are set in consideration of the response characteristics of the physiological index (SPR) to the stimulus, and th2> th1. When the SPR frequency is less than th1, the gain K = A is set. When the SPR frequency is greater than or equal to th1 and less than th2, the gain K = B is set. When the SPR frequency is greater than or equal to th2, the gain K = C is set. The values A, B, and C are set by experiments and the like, and are values from 0 to 1, and A> B> C.

In Equation (3), AR (t _s ) is an AR value for the applied wakefulness stimulus, and is calculated in the same manner as AR (t). τ is a gain for determining the degree of reflection of the AR value for the arousal stimulus. The gain τ is a value of 0 or more and less than 1, and is set as a two-stage value based on the SPR frequency. It should be noted that there is no limitation on how many stages of gain τ are set.

  If the SPR frequency is high when an arousal stimulus is applied to the driver, it can be understood that the arousal state is increased by the arousal stimulus (the effect of the arousal stimulus is present), and the driver's response to the stimulus is fast. In this case, it is necessary to largely reflect the effect of the arousal stimulus on the state estimation. In addition, it is necessary to quickly reflect the rapid state change of the driver in the state estimation so that the state estimation is not delayed with respect to the change of the driver state. Therefore, when the SPR frequency for the wakeful stimulus is high, the gain τ is increased, and the degree of reflection of the AR value for the wakeful stimulus is increased.

  Specifically, a threshold th3 for the SPR frequency is prepared. The threshold th3 is set in consideration of the response characteristic of a physiological index (SPR) to a stimulus, and th3> th2> th1. When the SPR frequency is less than th3, the gain τ = α is set. When the SPR frequency is greater than or equal to th3, the gain τ = β is set. α and β are set by experiment or the like, and are values from 0 to 1, where α <β.

  Next, processing executed by the ECU 8 using the sleepiness estimation model will be described in detail.

  The ECU 8 transmits a vibration generation signal indicating a fine parameter for reference stimulation to the vibration generator 2 at a constant cycle. This vibration generation signal is a signal for acquiring the drowsiness level of the driver, and the transmission interval may be arbitrarily set.

  The ECU receives an amplification detection signal from the amplifier 5 every time a fine vibration for reference stimulation is generated. Subsequently, the ECU 8 filters the SPA of the amplification detection signal, removes the SPL component from the SPA, and extracts the SPR component. Subsequently, the ECU 8 acquires (measures) the current SPR value by integrating the SPR component in the reaction time Tp set for the vibration generating device 2 according to the equation (1). Subsequently, the ECU 8 counts the number of times of coming out of the range of mean + 3 × σ to mean−3 × σ in the reaction time Tp of the SPR component, and sets it as the current SPR frequency. Subsequently, the ECU 8 calculates an average value (mean) and a standard deviation value σ in the reaction time Tp of the SPR component acquired this time in order to calculate the next SPR frequency.

  Subsequently, the ECU 8 determines whether the calculated SPR frequency is less than th1, th1 or more and less than th2, or th2 or more. At this time, the ECU 8 sets A as the gain K when the SPR frequency is less than th1, sets B as the gain K when it is equal to or greater than th1 and less than th2, and sets C as the gain K when equal to or greater than th2. Set. Further, the ECU 8 calculates the current AR value using the current SPR value and the previous SPR value according to the equation (2).

  When it is determined that the driver's arousal state is at a low level, the ECU 8 selects one or more devices from the vibration generator 2, the air conditioner 3, the speaker 6, and the display 7, and selects the selected device. Used to give the driver a wakeful stimulus. At this time, the ECU 8 may select a device based on the driver's arousal state (depth of sleepiness), or a predetermined time has elapsed to prevent the driver from getting used to a specific stimulus. Then, you may switch the apparatus which provides an arousal stimulus.

  Thereby, the state estimation system 1 can provide a driver with a plurality of types of arousal stimuli. For example, the state estimation system 1 has an immediate stimulus (first stimulus) in which the reaction time of the biological reaction is short (the rate of accepting somatic sensation is fast and the effect immediately appears), and the biological reaction is faster than this immediate effect stimulus. A slow-acting stimulus (second stimulus) that has a long reaction time (slow response to somatosensory and effects appear gradually) can be given to the driver. As described above, an example of the immediate effect stimulus is a vibration applied from the vibration generator 2, and an example of the delayed effect stimulus is a cold air applied by the air conditioner 3.

  Subsequently, the ECU 8 receives the amplification detection signal from the amplifier 5 and extracts the SPR component as in the case of the reference stimulus. Subsequently, the ECU 8 sets a reaction time (measurement time) used in the calculation using the equation (1) according to the applied arousal stimulus. Subsequently, the ECU 8 calculates (measures) the SPR value for the extracted SPR component by applying the extracted SPR component and the set reaction time to the equation (1). For example, the ECU 8 sets the reaction time Tp to calculate the SPR value corresponding to the wakefulness stimulus by the vibration generator 2, and sets the reaction time Tq to calculate the SPR value corresponding to the wakefulness stimulus by the air conditioner 3. To do. Subsequently, the ECU 8 obtains the SPR frequency in the reaction time and determines whether or not the SPR frequency is less than th3. Subsequently, the ECU 8 sets α as the gain τ when the SPR frequency is less than th3, and sets β as the gain τ when it is greater than th3.

  In addition, the ECU 8 calculates the AR value for the arousal stimulus by applying the SPR value for the arousal stimulus and the SPR value for the previous reference stimulus to Equation (2). Then, the ECU 8 calculates the current sleepiness level by applying the current AR value, the gain K, the AR value with respect to the arousal stimulus, the gain τ, and the previous sleepiness level to the sleepiness estimation model represented by Expression (3). . Note that if the arousal stimulus is not applied between the previous and previous reference stimuli, the AR value for the arousal stimulus is set to 0, and the current sleepiness level is calculated.

  Each time the ECU 8 obtains the current sleepiness level, the ECU 8 determines whether or not the sleepiness level is equal to or less than the arousal stimulus imparting level. The arousal stimulus imparting level is a threshold value for determining whether or not the level of the arousal state is low enough to cause the driver to interfere with driving even a little (a level where sleepiness is strong). For example, 4 (D2 ) Value is set. When the ECU 8 determines that the drowsiness level this time is equal to or less than the arousal stimulus imparting level, the ECU 8 imparts the awake stimulus to the driver.

  For example, the ECU 8 generates a voice message and an image for alerting the driver according to the sleepiness level, transmits a voice signal composed of the voice data to the speaker 6, and displays an image signal composed of the image data. 7 to send. Further, the ECU 8 sets a vibration generation signal for generating a relatively strong vibration and transmits it to the vibration generator 2 in order to improve the awakening state for the driver according to the sleepiness level. Further, the ECU 8 generates an air conditioning control signal for applying cool air for improving the driver's arousal state according to the sleepiness level, and transmits the air conditioning control signal to the air conditioner 3.

  At this time, the ECU 8 decreases the drowsiness level from 4 (the drowsiness increases), the image, the voice, the vibration, the image that gives the cold air step by step, the voice message, the vibration generation signal, An air conditioning control signal is generated. At this time, the ECU 8 may perform control to give all images, sound, vibration, and cold air, or may perform control to give one or more stimuli from among a plurality of arousal stimuli according to the drowsiness level.

  Next, the operation of the state estimation system 1 will be described with reference to FIG.

  The electrodermal activity sensor 4 detects sweating from the palm of the driver and transmits the detection signal to the amplifier 5. The amplifier 5 amplifies the detection signal from the electrodermal activity sensor 4 and transmits the amplified detection signal to the ECU 8.

  The ECU 8 sets a vibration generation signal for generating a fine vibration for reference stimulation at regular intervals and transmits it to the vibration generator 2. When this vibration generation signal is received, the vibration generator 2 generates a fine vibration for reference stimulation.

  Each time a reference stimulus is applied, the ECU 8 sets a reaction time (measurement time) corresponding to the reference stimulus. Subsequently, the ECU 8 obtains the SPR value for the current reference stimulus by the equation (1) based on the amplification detection signal from the amplifier 5 and the set reaction time, and the average value and standard deviation of the previous SPR component. The SPR frequency is obtained based on the value. Subsequently, the ECU 8 calculates the current AR value from the SPR value with respect to the current reference stimulus and the SPR value with respect to the previous reference stimulus by the equation (2). Further, the ECU 8 sets the gain K based on the SPR frequency for the current reference stimulus. At this time, the ECU 8 obtains the average value and the standard deviation value of the SPR component for the next time.

  In particular, when a wake-up stimulus is applied, the ECU 8 sets a reaction time (measurement time) corresponding to the wake-up stimulus. Subsequently, the ECU 8 obtains the SPR value for the wakefulness stimulus by the equation (1) based on the amplification detection signal from the amplifier 5 and the set reaction time, and calculates the average value and the standard deviation value of the previous SPR component. The SPR frequency is obtained as a reference. Subsequently, the ECU 8 calculates an AR value for the wakefulness stimulus from the SPR value for the wakefulness stimulus and the SPR value for the previous reference stimulus by the equation (2). Further, the ECU 8 sets the gain τ based on the SPR frequency for the arousal stimulus.

  Every time the AR value for the current reference stimulus is obtained, the ECU 8 uses the sleepiness estimation model represented by Equation (3) to determine the AR value for the current reference stimulus and the AR value for the arousal stimulus (if no arousal stimulus is applied). 0), the current sleepiness level is calculated from the gain K, the gain τ, and the previous sleepiness level. As a result, the sleepiness level of this time in which the degree of addition of the AR value is adjusted according to the SPR frequency is obtained. In particular, when the wakefulness stimulus is applied immediately before, the effect of the wakefulness stimulus on the current sleepiness level is obtained. Is also reflected.

  Subsequently, the ECU 8 determines whether or not the current drowsiness level is equal to or lower than the awakening stimulus imparting level. Then, when it is determined that the sleepiness level this time is equal to or less than the arousal stimulus imparting level (when the arousal state is low enough to cause any trouble in driving), the ECU 8 imparts the arousal stimulus to the driver. Specifically, the ECU 8 selects one or more devices from the vibration generating device 2, the air conditioner 3, the speaker 6, and the display 7, and sends a predetermined signal (audio signal, image signal, vibration generating signal) to the selected device. Or an air conditioning control signal). The ECU 8 repeats the above processing at regular intervals.

  When the speaker 6 receives the audio signal, the speaker 6 outputs a warning message according to the audio signal. When the display 7 receives the image signal, the display 7 displays a warning image according to the image signal. When receiving the vibration generation signal, the vibration generator 2 generates strong vibration. When the air conditioner 3 receives the air conditioning control signal, the air conditioner 3 sends cool air to the interior space. These arousal stimuli increase the driver's arousal state and increase the attention to driving.

  As described above, according to the present embodiment, by setting the reaction time (measurement time) of the biological reaction for each reference stimulus or awakening stimulus, the biological reaction for each stimulus, that is, the SPR value can be accurately measured, As a result, the drowsiness level (wakefulness state) can be accurately estimated. For example, as shown in FIG. 3, the response time Tp is set for the immediate effect stimulus and the reaction time Tq is set for the delayed effect stimulus, so that the SPR value for the SPR component Ep and the SPR value for the SPR component Eq are set. Since the value can be measured with high accuracy, the driver's arousal state can be estimated with high accuracy based on the measurement result.

  Such an effect will be further described with reference to FIG. FIG. 4 shows a case where the sleepiness level is estimated using a method using Equation (1) (hereinafter also referred to as “integration method”), and the maximum value and the minimum value of the SPR component at a predetermined time, which have existed conventionally. Is a graph compared with a case where a method of calculating a difference between these two values is extracted (hereinafter referred to as “Max-Min method”). The vertical axis and horizontal axis of the graph shown in FIG. 4 indicate the sleepiness level and time, respectively.

  The solid line Ex in FIG. 4 is the transition of sleepiness level acquired based on the SPR value measured using the integration method, and the broken line Ey is the transition of sleepiness level when the Max-Min method is used. An alternate long and short dash line Eo is a transition of the sensory evaluation value of the drowsiness level, and serves as a reference when comparing the integration method and the Max-Min method. Here, the sensory evaluation value is obtained by objectively evaluating a change in facial expression based on an image obtained by imaging the driver's face. The sleepiness levels Ex, Ey, Eo are results obtained from the actual vehicle test. As shown in FIG. 4, especially in the section I, the degree of deviation from the sensory evaluation value is smaller when the integration method is used than when the Max-Min method is used. That is, the drowsiness level can be accurately estimated by using the integration method. This is because the SPR value used when acquiring the sleepiness level is accurately measured.

  Such a difference in effect will be described with reference to FIG. FIG. 5 is a diagram comparing the case where two types of SPR values are measured using the integration method and the case where measurement is performed using the Max-Min method. The latency Ta, reaction times Tp, Tq, and SPR components Ep, Eq are the same as those shown in FIG.

  In the integration method used in the present embodiment, the reaction time is set according to the applied stimulus, and the SPR value is acquired based on the SPR component measured during the reaction time. That is, by using the integration method, it is possible to acquire biological information indicating a biological reaction for one cycle for any stimulus. For example, the SPR component Ep measured during the reaction time Tp is measured for the wakefulness stimulus (immediate effect stimulus) by the vibration generator 2, and for the wakefulness stimulus (slow effect stimulus) by the air conditioner 3, The SPR component Eq measured during the reaction time Tq is measured.

  Further, in the integration method, not only the amplitude of the SRR component (for example, the amplitude Ap for the SPR component Ep, the amplitude Aq for the SPR component Eq) but also an SPR value that takes into account the temporal transition of the sleepiness level is derived (measured). Since the time transition of the drowsiness level indicates a change in the state of the driver, the state change of the driver that changes with the transition of the time can be estimated with higher accuracy by considering the time transition of the sleepiness level.

  Therefore, it is possible to evaluate the arousal effect equally for arousal stimuli that differ in the appearance of biological responses, such as immediate-acting stimuli and delayed-acting stimuli. Moreover, the driver's arousal state can be estimated with higher accuracy by acquiring the SPR value measured using the integration method and applying the SPR value to the sleepiness estimation model.

  On the other hand, in the conventional Max-Min method, a fixed measurement time is used regardless of the type of stimulation. For example, when the measurement time is set to Tp in the example of FIG. 5, the amplitude Ap can be obtained for the SPR component Ep, but the amplitude Aq ′ is obtained for the SPR component Eq instead of the amplitude Aq that should be originally obtained. Is done. As a result, it is not possible to accurately measure the SPR value for the delayed-effect stimulus by the air conditioner 3, and the estimation accuracy of the drowsiness level is also lowered. In the Max-Min method, only the amplitude (Ap, Aq ′) of the SPR component is considered, and the time transition of the sleepiness level is not considered. Therefore, the estimation accuracy of the sleepiness level is lowered.

  In addition, according to the present embodiment, the sleepiness estimation model represented by the expression (3) is constructed, the gain for the AR value is provided, and the degree of reflection of the AR value on the sleepiness level can be adjusted. The drowsiness level can be estimated with high accuracy.

  Further, according to the present embodiment, by setting the gain based on the SPR frequency, it is possible to obtain the degree of reflection of the AR value according to the driver's state that changes from moment to moment, and to further improve the state estimation accuracy. be able to. Further, in the present embodiment, since a fast reaction index called SPR frequency is used, the driver's state can be quickly reflected in the gain (and the driver's state can be quickly reflected in the state estimation). State estimation accuracy can be further improved.

  In addition, according to the present embodiment, a slow reaction index called an AR value is used, but a change in the state of the driver is quickly and accurately reflected by determining the degree of consideration of the AR value based on a gain based on a fast reaction index called an SPR frequency. Can be estimated.

  As mentioned above, although embodiment which concerns on this invention was described, this invention is implemented in various forms, without being limited to the said embodiment.

  For example, in the above-described embodiment, the human state estimation apparatus according to the present invention is applied to the state estimation system 1, but it is also possible to apply the apparatus to other objects. For example, the device according to the present invention may be applied to a device that estimates the state of various people such as other vehicle drivers, various plant supervisors, night workers, etc. You may apply to the apparatus which estimates other states, such as (impression, irritation, boredom), a fatigue state, concentration, attention, and stress.

  Moreover, in the said embodiment, although the vibration generated from the seat was utilized as a reference stimulus given to a person, the kind of reference stimulus is not limited. For example, other physical stimuli such as sound and light may be used as reference stimuli, and vibration, light, and sound that are constantly emitted from the device may be used. Moreover, the physical stimulus converted from the stimulus received from an external environment may be sufficient, for example, the sound heard by a driver | operator may be detected and the driver | operator may be given the mechanical vibration converted from the detected sound. The stimulus received from the external environment is, for example, road noise generated from a tire. When the state estimation system 1 uses a plurality of types of reference stimuli, the state estimation system 1 stores in advance the reaction time (measurement time) set for each reference stimulus, and sets the reaction time corresponding to the applied reference stimulus. To derive the SPR value and SPR frequency.

  In the above embodiment, the reference stimulus is given at a constant cycle. However, the reference stimulus may be given at various timings. For example, the reference stimulus may be given at a random cycle, or the state of the driver is estimated. A reference stimulus may be applied when needed (eg, before each event is expected to occur). This event is an environment in which the driver is placed during driving, for example, traffic congestion that induces the driver's sleepiness, traveling on a highway.

  Moreover, in the said embodiment, although electrodermal activity (EDA), especially skin potential reaction (SPR) were utilized as a physiological index, various physiological indices can be utilized. For example, physiological indices (such as blinks) based on electrooculogram (EOG (Electrooculogram)), physiological indices (such as α waves and β waves) based on electroencephalograms (EEG (Electroencephalogram)), physiological indices based on electrocardiograms (ECG (Electrocardiogram)) Heart rate, etc.), physiological indices based on electromyogram (EMG) can be used. In addition, other physiological indicators such as skin resistance level (SRL), skin resistance response (SRR), and skin potential level (SPL) may be used in the electrodermal activity. Furthermore, the state may be estimated by combining a plurality of physiological indexes, instead of estimating the state using only one physiological index.

  In the above embodiment, the AR value is applied as the first index value, and the SPR frequency is applied as the second index value for setting the gain. However, the present invention is applicable to combinations other than the combination of the AR value and the SPR frequency. Is possible. For example, a combination of heart rate fluctuation LF as the first index value and a heart rate as the second index value, a combination of blink duration as the first index value and the number of blinks as the second index value, and heart rate fluctuation LF as the first index value And the combination of the number of blinks can be used as the second index value.

  Further, in the above embodiment, when the drowsiness level becomes a level that hinders driving, it is configured to be alerted by image display, sound output, vibration, etc., but attention should be paid by other means such as an alarm buzzer and smell. It is good also as a structure to arouse. In addition, when the condition becomes a level that hinders driving, it is safer by changing the control timing and control threshold of the driving support system (for example, pre-crash safety system, adaptive cruise control system, lane keeping system). You may make it control by the vehicle side so that property may be improved. Alternatively, the human state estimation device according to the present invention may be configured to detect a human state and output the human state.

It is a figure which shows the structure of the state estimation system which concerns on this Embodiment. It is a graph which shows the example of the SPR component detected. It is a graph which shows two types of SPR components. It is the graph which compared the estimation result of the drowsiness level with the method concerning this embodiment, and the conventional method. It is the figure which compared the method which concerns on this embodiment, and the conventional method regarding the measurement of SPR value.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... State estimation system, 2 ... Vibration generator, 3 ... Air conditioner, 4 ... Electric skin activity sensor, 5 ... Amplifier, 6 ... Speaker, 7 ... Display, 8 ... ECU

Claims (5)

A human state estimation device for estimating the state of a subject,
A stimulus applying means for applying a stimulus to the subject;
Biological information acquisition means for acquiring biological information indicating the biological reaction of the subject to the stimulus applied by the stimulus applying means;
A measurement time setting means for setting a measurement time for measuring the biological reaction according to the type of stimulus given by the stimulus giving means;
A biological reaction measuring means for measuring the biological reaction based on the biological information acquired by the biological information acquiring means during the measurement time set by the measuring time setting means;
State estimation means for estimating the state of the subject based on the biological reaction measured by the biological reaction measurement means;
A human state estimation device comprising: The types of stimuli include a first stimulus and a second stimulus that has a slower biological response to the stimulus than the first stimulus.
The human state estimating apparatus according to claim 1, wherein: The biological reaction measuring means measures the biological reaction by integrating biological information acquired during the measurement time set by the measurement time setting means.
The human state estimation apparatus according to claim 1 or 2, characterized in that First index value acquisition means for acquiring a first index value corresponding to the state change amount of the subject based on the biological reaction measured by the biological reaction measurement means;
A gain setting means for setting a gain for determining a degree of consideration of the first index value when estimating the state of the subject,
The state estimation means defines a degree of addition of the first index value acquired by the first index value acquisition means according to the gain set by the gain setting means, and a first index value of the degree of addition according to the gain And the subject's state based on the past subject's state,
The human state estimation device according to any one of claims 1 to 3, wherein Based on the biological reaction measured by the biological reaction measuring means, further comprising a second index value acquiring means for acquiring a second index value having a larger change with respect to a change in the state of the subject than the first index value;
The gain setting means sets a gain based on the second index value acquired by the second index value acquisition means;
The human state estimation device according to claim 4, wherein

Images