WO2017085756A1 - Dispositif, procédé et programme de traitement d'informations - Google Patents

Dispositif, procédé et programme de traitement d'informations Download PDF

Info

Publication number
WO2017085756A1
WO2017085756A1 PCT/JP2015/082081 JP2015082081W WO2017085756A1 WO 2017085756 A1 WO2017085756 A1 WO 2017085756A1 JP 2015082081 W JP2015082081 W JP 2015082081W WO 2017085756 A1 WO2017085756 A1 WO 2017085756A1
Authority
WO
WIPO (PCT)
Prior art keywords
angular velocity
zero point
information
sensor
acceleration
Prior art date
Application number
PCT/JP2015/082081
Other languages
English (en)
Japanese (ja)
Inventor
陽介 千田
崇尚 杉本
弘志 根來
Original Assignee
富士通株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 富士通株式会社 filed Critical 富士通株式会社
Priority to PCT/JP2015/082081 priority Critical patent/WO2017085756A1/fr
Publication of WO2017085756A1 publication Critical patent/WO2017085756A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C19/00—Gyroscopes; Turn-sensitive devices using vibrating masses; Turn-sensitive devices without moving masses; Measuring angular rate using gyroscopic effects
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
    • G01P21/00—Testing or calibrating of apparatus or devices covered by the preceding groups

Definitions

  • the present invention relates to an information processing apparatus, method, and program.
  • the object whose position is to be estimated is a pedestrian
  • PDR pedestrian Dead-Recking
  • the walking timing and the stride are estimated for each step using an acceleration sensor among the motion sensors 20 built in the terminal device 10, and further a pedestrian using a geomagnetic sensor.
  • the relative position (current position) from the initial position is estimated by calculating the direction in which the pedestrian travels and the distance walked by integrating these pieces of information. Is done.
  • Patent Document 1 the straight traveling / turning motion is determined by “comparison with a pattern in a walking motion given in advance”. Therefore, in the method of Patent Document 1, it is necessary to have a template for each walking motion pattern, and it is necessary to match the acquired data with each template. For this reason, memory consumption and calculation load are high. Similarly, Patent Document 2 consumes high computer resources.
  • the straight traveling state is determined based on whether or not the value obtained by integrating the output values of the angular velocity sensors in the two-step section of the pedestrian holding the terminal device simply exceeds the threshold value. For this reason, if the zero-point deviation amount of the output value of the angular velocity sensor is large, the integral value may exceed the threshold even in a state of actually going straight, and it may not be determined that the vehicle is moving straight, and correct determination may be difficult.
  • an object of the present invention is to accurately estimate the amount of zero point deviation of an angular velocity sensor during walking.
  • An information processing apparatus includes a setting unit that estimates a zero point deviation amount and subtracts the zero point deviation amount from an output value of the angular velocity sensor to set a zero point of the angular velocity sensor.
  • shift amount The figure for demonstrating rotation of an object.
  • the flowchart which shows an example of the variable initialization process for turbulence measurement functions of one Example The flowchart which shows an example of the variable update process for turbulence measurement functions of one Example.
  • the integrated value of the angular velocity sensor is an angle
  • the integrated value of the output is zero (zero) in the case of periodic movement in which the postures of the start point and the end point are equal.
  • the average of the angular velocity sensors in one period of such movement is the zero point deviation amount.
  • an average of a sufficiently long time for one cycle may be taken.
  • offset amount hereinafter also referred to as “offset amount” or “offset”
  • offset amount is applied to pedestrian autonomous navigation.
  • One example of a periodic movement that can be expected to have the same starting and ending postures under the premise is a two-step section during straight progress.
  • the zero point deviation amount can be calculated from the average value of the output values of the angular velocity sensor in the straight traveling state.
  • the terminal device 10 As a method for distinguishing between the straight traveling state and the non-straight traveling state, the terminal device 10 is kept stationary for a certain time after the terminal device 10 is started, and the average value of the output values of the angular velocity sensor or the like is used as the zero point deviation amount. It is possible. This method is often used in dedicated equipment such as robots. However, if this method is to be applied to a general product, the user must be forced to place the terminal device 10 in a stationary state for a certain period of time, and there is a possibility of giving up the commercial value. Furthermore, since the zero point deviation amount has a characteristic that changes with time, the zero point deviation amount calculated as time passes may not be accurate.
  • the zero point deviation can be estimated at any time.
  • One method is to apply a low-pass filter with a large time constant to the output of the angular velocity sensor, and consider this as the zero point deviation amount.
  • the low-pass filter since the low-pass filter only smoothes the input information in the time direction, when turning, information that is generated by turning is also smoothed and accumulated. Even if the turn is completed, the information that the turn has been made for a certain time remains, so that a new problem occurs in which the reaction is reversed.
  • the change amount of the acceleration sensor is small, it is considered that the rotational motion is not performed, and there is a method in which the average value of the angular velocity sensor during that time is set as the zero point deviation amount.
  • the angular velocity information and the acceleration information are independent, but in reality, it is used that there is almost no movement in which only one information is generated.
  • this method has a problem that it is difficult to set a threshold value for distinguishing a stationary state. For example, assuming a stationary state while walking as in the present embodiment, if the threshold value is strict, it is almost not determined to be a stationary state and zero point deviation calibration is not performed. However, if it is loosened, the accuracy of the estimated value of the zero point deviation amount is lowered.
  • the present embodiment provides an information processing apparatus capable of accurately estimating the zero point deviation amount of the angular velocity sensor during walking and improving the accuracy of the walking path estimation and the self-position estimation.
  • the terminal device 10 is an example of an information processing device.
  • the information processing device is not limited to this, but is a smartphone, a video camera, a digital camera, a PDA (Personal Digital Assistant), a mobile phone, a portable music playback device, a portable video processing device, a portable game device, an HMD (Head Mount Display). ), Wearable terminal devices such as a watch-type terminal device, and other electronic devices.
  • the terminal device 10 includes a CPU 11, a memory 12, an input / output I / F 13, a touch panel 14, a display 15, a communication I / F 16, an acceleration sensor 21, an angular velocity sensor 22, and a geomagnetic sensor 23.
  • the acceleration sensor 21, the angular velocity sensor 22, and the geomagnetic sensor 23 are examples of motion sensors that detect a walking motion of a person holding the terminal device 10.
  • the acceleration sensor 21, the angular velocity sensor 22, and the geomagnetic sensor 23 are built in the terminal device 10, but if the acceleration sensor 21 and the angular velocity sensor 22 are built in or externally attached to the terminal device 10, the geomagnetism.
  • the sensor 23 may not be incorporated.
  • the geomagnetic sensor 23 may not be included in the motion sensor 20.
  • the acceleration sensor 21 is included in the motion sensor 20 in the case where control is performed using a predetermined two-step time instead of detecting a “predetermined one cycle” described later using an acceleration waveform. It does not have to be.
  • the CPU 11 controls the terminal device 10 according to a program stored in the memory 12.
  • the memory 12 is, for example, a semiconductor memory, and stores a program executed by the CPU 11, data referred to by the CPU 11, data acquired as a result of processing executed by the CPU 11, and the like.
  • At least a part of the zero point correction processing program and data stored in the recording medium 17 may be copied to the memory 12 as necessary, or the acquired data is copied from the memory 12 to the recording medium 17 as necessary. May be.
  • the recording medium 17 is a non-volatile recording medium such as a flash memory.
  • the input / output I / F 13 is an interface for accepting input of information from a pedestrian holding the terminal device 10 and providing information to the pedestrian.
  • the input / output I / F 13 receives input of information from an input device such as a touch panel 14, a keyboard, a button, or a pointing device.
  • the input / output I / F 13 displays various information to a pedestrian using an output device such as a display 15.
  • an instruction to start data acquisition such as an angular velocity sensor value, an acceleration sensor value, and a geomagnetic sensor value is input / output I / O
  • An instruction to end data acquisition may be input to the input / output I / F 13 when input to F13 and exit from the target area.
  • Data acquisition of various sensor values may be automatically started in conjunction with activation of the terminal device 10.
  • the CPU 11 controls the angular velocity sensor 22 and the acceleration sensor 21 from when it is instructed to start data acquisition until it is instructed to end, for example, periodically measures the angular velocity and acceleration, and stores the results in the memory 12. To store.
  • the communication I / F 16 is an interface that is connected to a network and communicates with other devices such as server devices.
  • the acceleration sensor 21 measures the acceleration of the terminal device 10.
  • the acceleration sensor 21 measures accelerations in the axial directions of the X axis, the Y axis, and the Z axis with respect to the orientation of the terminal device 10.
  • the angular velocity sensor 22 measures the angular velocity of the terminal device 10.
  • the angular velocity sensor 22 measures angular velocities around the X axis, the Y axis, and the Z axis with respect to the orientation of the terminal device 10.
  • the geomagnetic sensor 23 detects geomagnetism and measures the absolute direction (hereinafter also referred to as “geomagnetic vector”) based on it.
  • the geomagnetic vector changes greatly at a slight distance of several meters, such as near a speaker, a reinforcing bar, or a large power line. In such a place where the magnetic field is disturbed, the output value of the geomagnetic sensor 23 may not accurately measure the direction of the magnetic field.
  • the terminal device 10 has the functions of an acquisition unit 110, a counting unit 111, a calculation unit 112, a setting unit 113, and a recording unit 114.
  • the acquisition unit 110 acquires the sensor value output by the motion sensor.
  • the acquisition unit 110 acquires the sensor value detected by the acceleration sensor 21, and the recording unit 114 records X-axis, Y-axis, and Z-axis acceleration information in the acceleration information table 116 based on the acquired sensor value.
  • the acquisition unit 110 acquires the sensor value detected by the angular velocity sensor 22, and the recording unit 114 records the X-axis, Y-axis, and Z-axis angular velocity information in the angular velocity information table 117 based on the acquired sensor value.
  • the acquisition unit 110 may acquire the sensor value detected by the geomagnetic sensor 23. In that case, the recording unit 114 records a geomagnetic vector (absolute direction information) in the geomagnetic information table 118 based on the acquired sensor value.
  • the counting unit 111 counts the number of steps of the user using the acceleration waveform based on the acquired acceleration information.
  • the calculation unit 112 uses the acquired angular velocity to calculate information related to the first angular velocity in a predetermined cycle and information related to the second angular velocity in a cycle following the predetermined cycle.
  • the setting unit 113 sets the zero point deviation amount according to the angular velocity.
  • the zero point of the angular velocity sensor 22 is set by subtracting from the output value of the angular velocity sensor 22 (see FIG. 2).
  • the recording unit 114 records a zero point correction processing program 119 for correcting the sensor value of the angular velocity sensor 22 and a control program 120 for controlling the entire terminal device.
  • the terminal device 10 realizes control of the entire device through processing that the control program 120 causes the CPU 11 to execute.
  • the terminal device 10 realizes the functions of the counting unit 111, the calculation unit 112, and the setting unit 113 by processing that the CPU 11 executes by the zero point correction processing program 119.
  • FIG. 5A shows (1) from the initial state, (2) 90 degrees around the x axis, (3) 90 degrees around the y axis, and (4) around the x axis.
  • FIG. 5 is a diagram of ⁇ 90 °, (5) when rotated by ⁇ 90 ° around the z axis.
  • the start point (1) and the end point (5) are in the same posture.
  • the amount of movement around each axis is 0 °, which is the sum of (2) and (4) around the x-axis, but the remaining axis is 90 ° with the y-axis being (3) and the z-axis is (5) ⁇ 90 ° and not 0 °.
  • FIG. 5 (a) or FIG. 5 (b) which is half of the symmetrical movement. That is, FIG. 5A shows the rotation of the half cycle section (first half) when two steps are taken as one cycle, and FIG. 5B shows the rotation of the half cycle section (second half).
  • Each angular movement amount of the angular velocity sensor becomes zero when two steps are taken as one set, and the total (average value) of angular velocities in one cycle section becomes zero.
  • two steps are set as one set, and two steps of the pedestrian are set as “a predetermined cycle”.
  • the predetermined one cycle is not limited to two steps.
  • a period in which even steps such as four steps, six steps, and eight steps are counted may be a predetermined one cycle, or an odd number such as three steps and five steps.
  • the period during which steps are counted may be a predetermined cycle.
  • the terminal device 10 can set a period in which a predetermined number of steps of two or more steps is counted when the “predetermined one cycle” is determined.
  • the average angular velocity in one cycle section (two steps) in the straight traveling state is not non-zero. Since the theoretical average is zero, this value is the zero point deviation amount. Since the zero point deviation amount is unknown in the initial state, it is impossible to identify whether the pedestrian is turning or the zero point deviation is generated only by the information that the average value is non-zero.
  • the average difference between the different two-step sections is small, it can be determined whether the vehicle is going straight ahead. Actually, the difference is small even when turning with a constant curvature, but assuming that the operating environment is an artificial structure that GPS does not reach, it can be assumed that it will hardly move for a long time with a constant curvature. That is, it is possible to estimate that the pedestrian is going straight by making the average difference of the angular velocities of different one-cycle sections smaller than a certain threshold value, and the average value can be used as the sensor zero point deviation amount.
  • the counting unit 111 detects a two-step section that is the number of steps in one cycle of the pedestrian based on the sensor value of the acceleration sensor 21 in S101.
  • a known technique can be used for the detection method, and is omitted here.
  • the calculation unit 112 calculates an average value of the sensor values of the angular velocity sensor 22 in the section.
  • the calculation unit 112 detects the sensor of the angular velocity sensor 22 in the current and previous two-step sections. The difference between the average values is calculated and compared, and it is determined in S104 whether the difference is larger than a predetermined first threshold.
  • the average value of the two-step section calculated last time is the average value of the sensor values of the angular speed sensor 22 of the current two-step section obtained in S102. What is necessary is just to record as an average value of the sensor value of the sensor 22.
  • the calculation unit 112 determines that the pedestrian is going straight, and the sensor value of the angular velocity sensor 22 in the two-step section calculated in S102.
  • the average value is determined as the zero point deviation amount of the sensor.
  • the setting unit 113 determines angular velocity information by subtracting the zero point deviation amount from the sensor value of the angular velocity sensor 22 in accordance with pedestrian autonomous navigation. In this way, correction of the zero point of the angular velocity sensor 22 is executed. Since the average value of the sensor values of the angular velocity sensor 22 is somewhat dispersed, a plurality of average values obtained every time S105 is entered may be further averaged or smoothed using an appropriate low-pass filter. Good.
  • the calculation unit 112 determines that the pedestrian is not going straight, and does not update the zero point deviation amount (S107).
  • the average value of the sensor values of the angular velocity sensor 22 when the pedestrian is determined to be in the straight traveling state is zero.
  • the amount of deviation is determined, and the sensor value when it is determined that the vehicle is in the non-straight running state is discarded.
  • the zero point shift amount of the angular velocity sensor 22 during walking can be estimated with high accuracy.
  • the accuracy of the estimated zero deviation can be increased if the zero deviation is not updated.
  • the first is when large acceleration occurs. Generation of an acceleration greater than normal means that normal walking is not performed, and there is a high possibility that the assumption that the posture before and after one cycle section does not change is broken. Therefore, in the present embodiment, when a large acceleration occurs, the zero point deviation amount is not estimated and updated.
  • the second is when the change in the geomagnetic vector is large.
  • the present embodiment assumes operation in an environment where geomagnetism is unstable. This only means that the geomagnetic vector does not indicate magnetic north, and there is no significant variation in the geomagnetic vector in a local range of about 10 m. Therefore, when the average value of the geomagnetic vector before and after the two-step section is changing greatly, the zero point deviation amount is not corrected. For example, as in area A shown in FIG. 9, there is a place where the geomagnetic vector is largely switched at a slight distance of several meters such as near a reinforcing bar or a large power line in some artifacts. When passing there, the geomagnetic vector changes despite going straight. When passing through such a location, the zero point deviation amount is not estimated or updated even in the straight traveling state. However, even if such control is performed, there are many chances for the estimation update, so that there is no particular problem.
  • the zero point deviation amount is not updated, thereby increasing the estimation accuracy of the zero point deviation amount of the angular velocity sensor 22 during walking. Can be increased.
  • This processing is started, and after the processing of S101 and S102, the calculation unit 112 calculates the average value of the sensor values of the geomagnetic sensor 23 in the two-step section (S111), and averages the sensor values of the acceleration sensor 21 in the two-step section. A value is calculated (S112).
  • the calculation unit 112 determines that the pedestrian is in a straight traveling state. It is determined that there is a high possibility, and the process proceeds to S113.
  • the calculation unit 112 calculates a difference between the average values of the sensor values of the geomagnetic sensor 23 in the current and previous two-step sections, and in S114, the difference is larger than a predetermined second threshold value. To compare. In S114, when it is determined that the difference is smaller than the second threshold, the calculation unit 112 determines that there is a high possibility that the pedestrian is in a straight traveling state, and proceeds to S115.
  • the calculation unit 112 calculates the difference between the average values of the sensor values of the acceleration sensor 21 in the current and previous two-step sections, and in S116, the difference is larger than a predetermined third threshold value. To compare. In S116, when it is determined that the difference is smaller than the third threshold, the calculation unit 112 determines that there is a high possibility that the pedestrian is in a straight traveling state, and averages the sensor values of the angular velocity sensor 22 calculated in S102. The value is determined as the zero point deviation amount of the angular velocity sensor value.
  • the calculation unit 112 determines that the pedestrian is in a non-straight running state, and does not update the zero point deviation amount (S107).
  • the terminal device 10 when it is determined that the pedestrian is in the straight traveling state, the average value of the sensor values of the angular velocity sensor 22 is determined as the zero point deviation amount, and the non-straight traveling is performed. The sensor value when determined to be in the state is discarded. Thereby, the zero point shift amount of the angular velocity sensor 22 during walking can be estimated with high accuracy.
  • the sensor of the geomagnetic sensor 23 is used.
  • the value is larger than the second threshold value or when the sensor value of the acceleration sensor 21 is larger than the third threshold value, it is determined that the pedestrian is not in a straight traveling state.
  • the accuracy of the estimated value of the zero point deviation amount of the angular velocity sensor during walking can be further increased.
  • the zero point deviation amount of the angular velocity sensor 22 during walking is estimated with higher accuracy by a method different from the second embodiment.
  • straight travel determination may be performed using the past six steps. If only two sets of turns are used in the past six steps, that is, two sets of two steps, if the average value of the three sets is almost the same, it can be determined that the vehicle is going straight.
  • Fig. 12 shows a flowchart of straight ahead determination and zero point correction processing using 6 steps.
  • the difference from the first embodiment is that the processes of S117 and S118 are newly added to steps 101 to S107 of the straight-ahead determination and zero point correction process (FIG. 8) according to the first embodiment.
  • the calculation unit 112 compares the average value of the sensor values of the angular velocity sensor 22 in the current two-step section and the previous two-step section. If it is determined in S104 that the difference is greater than the first threshold value, the calculation unit 112 compares the average value of the sensor values of the angular velocity sensor 22 in the current two-step section and the previous two-step section in S117. In S118, it is determined whether the difference is larger than the first threshold.
  • the calculation unit 112 determines that the pedestrian is in a straight traveling state, and calculates the average value of the sensor values of the angular velocity sensor 22 calculated in S102 as a sensor. Is determined as the zero point deviation amount. Note that the average value of the sensor values of the angular velocity sensor 22 in the current two-step section, the previous two-step section, and the previous two-step section may be determined as the zero point deviation amount of the sensor. If it is determined in S118 that the difference is greater than the first threshold, the calculation unit 112 determines that the pedestrian is in a non-straight running state and does not update the zero point deviation amount (S107).
  • the past six steps determines whether the pedestrian travels straight from three walks.
  • the accuracy of the estimated value of the zero point deviation amount of the angular velocity sensor 22 can be further increased.
  • the straight-ahead determination and zero point correction process (FIG. 12) according to the third embodiment
  • the straight-ahead determination and zero point correction process (FIG. 8) according to the first embodiment
  • the straight-ahead determination and zero point correction processing (FIG. 10) according to the second embodiment may be expanded.
  • the straight traveling, stationary determination and zero point correction processing according to the fourth embodiment will be described with reference to FIG.
  • the zero point correction process is executed based on the straight traveling determination process.
  • zero point correction processing is performed including stillness determination.
  • the vehicle it is determined whether or not the vehicle is going straight when walking, and based on the average value of the sensor values of the angular velocity sensor 22 acquired when it is determined that the vehicle is in a straight traveling state, The zero point correction of the angular velocity sensor 22 was executed.
  • the owner is not always walking, and may stop or sit on a chair. In such a state, since the terminal device 10 is not moving, it is optimal for measuring the zero point deviation amount.
  • the utility can be improved by combining the stillness determination and the zero point correction process with the straight-ahead determination and the zero point correction process according to the first embodiment.
  • FIG. 13 shows a flowchart of stillness determination and zero point correction processing.
  • the counting unit 111 clears the number of steps W (S120), and determines whether the walking is an acceleration pattern (acceleration waveform) based on the acceleration sensor 21 (S121).
  • the counting unit 111 detects one step (S123) and confirms whether or not the second step is taken (S124).
  • the elapsed time is cleared (S128), and the zero point deviation amount is updated by the average value of the angular velocity sensor 22 through S102 to S106 as in FIG. .
  • the calculation part 112 determines whether there exists intense movement (S125).
  • the elapsed time is cleared (S129), and the process returns to S101.
  • the calculation unit 112 checks whether or not the state (S125) has passed for a certain time (S126). When the predetermined time has elapsed in S126, the calculation unit 112 clears the elapsed time (S127), proceeds to S105, and sets the average value of the sensor values of the angular velocity sensor 22 during that time as the zero point deviation amount. Note that the elapsed time clear in S127 to S129 is the one in which the timer is initialized in order to perform the next determination process in S126.
  • Pedestrian autonomous navigation has a relatively large amount of computation and memory consumption, and there are cases where there is no room for computer resources. In such cases, only a few tens of bytes of memory can be valuable.
  • the time taken for the two steps set in advance is set as one cycle, and the time taken for the two steps set in advance A two-step section may be detected.
  • S101 of the zero point correction process (FIG. 8) of the first embodiment S101 of the zero point correction process (FIG. 10) of the second embodiment, and S101 of the zero point correction process (FIG. 12) of the third embodiment.
  • the two-step section may be detected by the time required for two steps set in advance in S101 of the zero point correction process (FIG. 13) of the fourth embodiment.
  • FOG. 13 the zero point correction process
  • the calculation unit 112 clears an array DIM indicating N arrays in S200. This is an initialization process of the array DIM for S216 and S217 (averaging process) to “average further the average value of the sensor values of the angular velocity sensors 22 in a plurality of two-step sections”.
  • the memory consumption is about several hundred bytes.
  • Pedestrian autonomous navigation (PDR) is a variable for recording several Kbytes, for example, about 1 to 1.5 seconds.
  • the angular velocity sensor 22 has three axes of X, Y, and Z axes, in addition to this array DIM [], SUM (total value) and Ave [] (average value) described later are also three-dimensional. Here, it is shown in one dimension for simplicity.
  • the calculation unit 112 initializes the index Index for the averaging process (S217) (S201), and then sums the angular velocity sensor 22 for the current two steps (the current two steps) (integration result). And the variables Ave [0], Ave [1], Ave [2] for storing the average value of the angular velocity sensor values in the sections of the current two steps, the previous two steps, and the previous two steps. Is cleared (S202).
  • the calculation unit 112 initializes various variables for determining whether or not there is a strong movement in S270.
  • the “violent movement” shown in S125 of FIG. 13 is detected, the initialization process of S250 is required. Specific processing contents will be described later with reference to FIG.
  • the calculation unit 112 clears the step count W and the elapsed time cnt (S203).
  • the cnt is incremented for each loop in S204. Since this loop is executed at every sampling interval of the angular velocity sensor 22, this value is proportional to the elapsed time.
  • the calculation unit 112 executes a variable update process for detecting intense movement, similar to S250.
  • the calculation unit 112 increments cnt in S204, and then adds the sensor value of the angular velocity sensor 22 to the variable SUM in S205. Since the angular velocity sensor 22 has three detection axes as described above, the detection is actually performed for each axis.
  • the counting unit 111 calculates whether the user is walking (walking determination) or how many steps (step count measurement) from the information related to the sensor value of the acceleration sensor 21. Since this part only needs to utilize the technology of the existing pedometer, the description of the specific processing content is omitted.
  • the counting unit 111 compares in S208 whether the stored number of steps W is equal to the number of steps measured in S206. If these values are not equal (S208: No), the number of steps has not been increased, so the counting unit 111 does not perform processing (zero point correction) regarding updating of the offset (zero point deviation amount) (S224), Proceed to the next loop (processing after S260).
  • the counting unit 111 substitutes the number of steps for W (S209), and when the number of steps W is not an even number (S210: No), processing related to offset update (zero point correction) (S224), the process proceeds to the next loop (the process after S260).
  • step S209 stores the number of steps W for the next evaluation.
  • the calculation unit 112 performs a violent motion determination process (S270) for performing a zero point deviation amount update process every two steps during walking.
  • the counting unit 111 compares whether cnt is larger than an arbitrary value T, and if cnt is larger than an arbitrary value T, the processing of S270 is executed. When cnt is equal to or less than an arbitrary value T, the counting unit 111 does not perform the process related to offset update (zero point correction) (S224), and proceeds to the next loop (process after S260). Thereby, even when not walking, the zero point deviation amount updating process is performed at an appropriate interval. The processing content of S270 for confirming whether there is intense movement will be described later.
  • the calculation unit 112 compares the average value of the sensor values of the angular velocity sensor 22 in the current two-step section and the previous two-step section, and in S215, the current two-step section and the previous two-step section.
  • the average values of the sensor values of the angular velocity sensors 22 are compared, and it is determined whether each difference is larger than the first threshold value. This corresponds to the processing of S103, S104, S117, and S118 in FIG.
  • the calculation unit 112 since the variable Ave is three-dimensional, the calculation unit 112 performs the comparison for each axis, and if the difference is larger than a predetermined threshold even for one axis, the determination is “Yes” at each step. The process proceeds to S220.
  • the calculation unit 112 detects the sensor of the current two-step angular velocity sensor 22 in S216.
  • the average value Ave [0] of the values is stored in an array DIM [Index] indicating the average value of the angular velocities in the past that are relatively close.
  • the calculation unit 112 divides the addition value of the arrays DIM [0] to DIM [N ⁇ 1] by N, and substitutes the calculated average value of the array DIM for the zero point deviation amount (offset) (S217).
  • the setting unit 113 performs the calculation shown in FIG. 2 based on the zero point deviation obtained here. That is, the setting unit 113 subtracts the zero point deviation amount from the sensor value of the angular velocity sensor 22 to determine angular velocity information.
  • the calculation unit 112 increments the variable Index, and when the size of the variable Index is N or more, returns the variable Index to zero (S219). Thereafter, the calculation unit 112 clears the variables SUM and cnt for the next calculation regardless of whether or not the offset is updated (S220, S221). In addition, variables used for the violent motion determination process executed in S270 are initialized (S290), and the process returns to S260.
  • FIG.15, FIG.16, FIG.17, and FIG. 18 show the detailed flow of the process S250, S260, S270, and S290 for detecting the intense motion performed in FIG.
  • variable initialization In the variable initialization process in S250, as shown in FIG. 15, the calculation unit 112 clears MAG and GRV that store the information of the past gravity vector as the sensor value of the geomagnetic sensor 23 in the current section in S251. . Further, the calculation unit 112 clears Mg [] storing the past geomagnetic vector in S252, and ends this process.
  • both of these two variables and the ACC for storing information of the angular velocity sensor 22 described later are variables for storing the values of the three-axis sensor, they are actually three-dimensional variables and are processed for each axis.
  • variable update process In the variable update process of S260, as shown in FIG. 16, in S261, the calculation unit 112 adds the geomagnetic vector that is the sensor value of the geomagnetic sensor 23 to the variable MAG. Further, in S261, the calculation unit 112 stores the current sensor value of the acceleration sensor 21 in the array ACC [cnt]. Here, since cnt indicating the index of the array is incremented in S204, all of the values up to the two-step interval or cnt ⁇ T are stored, and this process ends.
  • the calculation unit 112 records the average value of the geomagnetic vector in the previous two-step section as the average value of the geomagnetic vector in the previous two-step section.
  • the calculation unit 112 records the average value of the geomagnetic vector in the current two-step section as the average value of the geomagnetic vector in the previous two-step section.
  • the calculation unit 112 records the average value Mg [0] of the geomagnetic vector in the current (current) two-step section from the variable MAG and the variable cnt of the geomagnetic vector.
  • the calculation unit 112 determines the average value Mg [0] of the geomagnetic vector in the current (current) two-step section and the average value Mg [2] of the two-step section from the previous four steps to the sixth step. ] Is significantly different from a predetermined threshold, it is determined that the robot is moving violently. In this case, in S281, the calculation unit 112 sets “1” indicating that there is a strong movement in the determination flag, and ends this process.
  • step S274 the calculation unit 112 calculates the average value Mg [0] of the geomagnetic vector in the current (current) two-step section and the average value Mg [2 of the two-step section from the previous four steps to the sixth step. ] Is less than or equal to a predetermined threshold value, it is determined that the geomagnetic vector has not changed so much, and the average value of accelerations ACC [0 to cnt] of the current (current) two-step section is calculated, Substitute in the gravity direction g (S275).
  • the calculation unit 112 takes an inner product (g ⁇ GRV) of the gravity direction g and the gravity GRV in the previous two-step section (two-step section from two steps before to four steps), and the inner product is less than a predetermined threshold value. It is determined whether it is larger (S276).
  • the inner product of the previous and current gravity directions means that G2 cos ⁇ is being calculated.
  • G is the magnitude of gravity, and its value is a constant value depending on the magnification of the acceleration sensor.
  • ⁇ is an angle formed by the gravity direction between the previous two-step section and the current two-step section.
  • the calculation unit 112 obtains the maximum value max of the difference between the average value of the acceleration sensor 21 and the gravity direction g in the current two-step section (S277), and the maximum value max falls within a predetermined allowable range. It is determined whether it exceeds (S278). In S278, vibration is always generated during walking, and a shocking component is removed, and it is confirmed whether the terminal is placed in a stationary state during non-walking.
  • the predetermined allowable range used in the determination process of S278 is large when walking (when it comes from a Yes path in S207 in FIG. 14), and is non-walking (Yes path in S222 in FIG. 14). It is desirable to keep it small).
  • variable initialization processing In the variable initialization process of S290, as shown in FIG. 18, the variable MAG indicating the geomagnetic vector is initialized, and this process ends.
  • FIG. 19 shows an example of actually measured data.
  • the horizontal axis in FIG. 19 represents time, and an example is given in which a transition is made from turning to straight traveling at time A and a transition from straight traveling to turning is made again at time B.
  • Three lines C1 to C3 shown in the uppermost graph show examples of sensor values of the acceleration sensor 21.
  • a line D is obtained by estimating the direction of gravity from here and converting it into acceleration in the vertical direction.
  • the counting unit 111 counts one vertical vibration of the line D.
  • the counting unit 111 can be realized by a pedometer.
  • E is marked with a marker every two steps, and a section F of the adjacent E becomes a “two-step section”.
  • G indicates an example of the sensor value of the angular velocity sensor 22.
  • sensor value data is output for the three axes of the X, Y, and Z axes.
  • Y-axis sensor value that substantially coincides with the vertical direction having the greatest influence is displayed with a line G.
  • the Y-axis and the vertical direction substantially coincide with each other because the measuring instrument is installed as such.
  • the average value of the sensor values G of the angular velocity sensor 22 in the two-step section F is H. This H is calculated for each step. For convenience, they are called h1, h2,.
  • the estimated zero point deviation amount K is updated.
  • the estimated zero point deviation amount does not decrease at a time but changes little by little.
  • the estimated offset amount is not updated in the L section.
  • 20A and 20B show the transition of the angle integral value (integral value of the sensor value (angle) of the angular velocity sensor 22) when the block on the square is made a round. This value is the traveling direction in FIG. In FIG.
  • the final integrated value is desirably ⁇ 360 ° for line I in FIG. 20A and 360 ° for line J in FIG. 20B.
  • the line I and the line J are finally close to desirable values as a result of the angle integral value when the zero point correction according to the present embodiment is performed.
  • the line K and the line L have jumped out of the graph from an early stage immediately after startup due to the transition of the angle integral value when the zero point correction according to the present embodiment is not applied.
  • the final estimated orientations of the line K and the line L were ⁇ 1780 ° for the line K in FIG. 20A and ⁇ 1180 ° for the line L in FIG. 20B. From the above, it can be seen from the results using measured data that the zero point deviation amount of the angular velocity sensor 22 during walking can be accurately estimated by correcting the zero value of the sensor value of the angular velocity sensor 22.
  • FIG. 21A to FIG. 21E show an example of the system configuration.
  • the operating environment of the program for executing the zero point correction according to the present embodiment may be the embedded device 10a shown in FIGS. 21A to 21C.
  • a pedestrian carries a built-in device 10a (which may be incorporated in a name card case, a belt buckle, a hat, etc.) having the function described using the terminal device 10 in the above embodiment.
  • the embedded device 10a estimates the zero point deviation amount by executing a program by the CPU 11, executes the zero point correction of the sensor value of the angular velocity sensor 22, and estimates the position of the pedestrian.
  • the position information indicating the estimated position of the pedestrian is transferred to the server device 30 by a wireless communication means such as Wi-Fi (registered trademark) or Bluetooth (registered trademark) or a log function.
  • the server device 30 can provide services such as centralized management of the location of each user using the transferred location information.
  • the present invention is also applied to a system that transfers position information to a single device such as the smartphone 40 without transferring the position information to the server device 30. it can.
  • the calculation for estimating the zero point deviation amount of the present embodiment is performed using the smartphone 10b or the mobile phone as one means of the embedded device, and is transferred to the server device 30 as shown in FIG. 21D.
  • the smartphone 10b may operate alone.
  • the angular velocity sensor 22, the acceleration sensor 21, the geomagnetic sensor 23, and the program do not need to be recorded in the same device, and the server device 30 logs or uploads the sensor information itself depending on the situation.
  • a method of applying the program in the server device 30 is also conceivable.
  • Terminal device 10
  • Memory 13
  • Touch panel 15
  • Display 16
  • Communication I / F DESCRIPTION OF SYMBOLS 21
  • Acceleration sensor 22
  • Angular velocity sensor 23
  • Geomagnetic sensor 30
  • Server apparatus 110 Acquisition part 111
  • Count part 112
  • Calculation part 113
  • Setting part 114 Recording part

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Navigation (AREA)
  • Gyroscopes (AREA)

Abstract

L'invention concerne un dispositif de traitement d'informations comprenant une unité d'acquisition pour acquérir une vitesse angulaire délivrée en sortie par un capteur de vitesse angulaire, une unité de calcul pour utiliser la vitesse angulaire acquise pour calculer des informations concernant une première vitesse angulaire dans une période prescrite et des informations concernant une seconde vitesse angulaire dans la période suivante après la période prescrite, et une unité de réglage pour, si la différence entre les informations calculées concernant la première vitesse angulaire et la seconde vitesse angulaire ne dépasse pas un premier seuil prédéfini, estimer une quantité de déviation du point zéro correspondant à la vitesse angulaire et régler le point zéro du capteur de vitesse angulaire par soustraction de la quantité de déviation du point zéro à partir des valeurs de sortie du capteur de vitesse angulaire.
PCT/JP2015/082081 2015-11-16 2015-11-16 Dispositif, procédé et programme de traitement d'informations WO2017085756A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2015/082081 WO2017085756A1 (fr) 2015-11-16 2015-11-16 Dispositif, procédé et programme de traitement d'informations

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2015/082081 WO2017085756A1 (fr) 2015-11-16 2015-11-16 Dispositif, procédé et programme de traitement d'informations

Publications (1)

Publication Number Publication Date
WO2017085756A1 true WO2017085756A1 (fr) 2017-05-26

Family

ID=58718545

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2015/082081 WO2017085756A1 (fr) 2015-11-16 2015-11-16 Dispositif, procédé et programme de traitement d'informations

Country Status (1)

Country Link
WO (1) WO2017085756A1 (fr)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08261761A (ja) * 1995-03-23 1996-10-11 Nippon Telegr & Teleph Corp <Ntt> ジャイロ計測におけるオフセットキャンセル方法
JPH10246642A (ja) * 1996-11-22 1998-09-14 Zexel:Kk 方位検出センサを使用して角速度補正を行うナビゲーション方法
JP2006071473A (ja) * 2004-09-02 2006-03-16 Alpine Electronics Inc 角速度センサのゼロ点誤差検出装置および方法
WO2007013216A1 (fr) * 2005-07-28 2007-02-01 Pioneer Corporation Dispositif de calcul d’azimut, méthode de calcul d’azimut, programme de calcul d’azimut et support d’enregistrement
JP2010038673A (ja) * 2008-08-04 2010-02-18 Smk Corp 動き検出ユニット
JP2010078567A (ja) * 2008-09-29 2010-04-08 Victor Co Of Japan Ltd 角速度センサ補正装置および角速度センサ補正方法
JP2011112500A (ja) * 2009-11-26 2011-06-09 Fujitsu Ltd センサ補正プログラム、センサ補正装置およびセンサ補正方法
JP2013152165A (ja) * 2012-01-25 2013-08-08 Fujitsu Ltd 検出装置、検出プログラム、及び検出方法
WO2014185027A1 (fr) * 2013-05-15 2014-11-20 旭化成株式会社 Dispositif et procédé d'estimation de décalage et programme
JP2015184160A (ja) * 2014-03-25 2015-10-22 セイコーエプソン株式会社 参照値生成方法、運動解析方法、参照値生成装置及びプログラム

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08261761A (ja) * 1995-03-23 1996-10-11 Nippon Telegr & Teleph Corp <Ntt> ジャイロ計測におけるオフセットキャンセル方法
JPH10246642A (ja) * 1996-11-22 1998-09-14 Zexel:Kk 方位検出センサを使用して角速度補正を行うナビゲーション方法
JP2006071473A (ja) * 2004-09-02 2006-03-16 Alpine Electronics Inc 角速度センサのゼロ点誤差検出装置および方法
WO2007013216A1 (fr) * 2005-07-28 2007-02-01 Pioneer Corporation Dispositif de calcul d’azimut, méthode de calcul d’azimut, programme de calcul d’azimut et support d’enregistrement
JP2010038673A (ja) * 2008-08-04 2010-02-18 Smk Corp 動き検出ユニット
JP2010078567A (ja) * 2008-09-29 2010-04-08 Victor Co Of Japan Ltd 角速度センサ補正装置および角速度センサ補正方法
JP2011112500A (ja) * 2009-11-26 2011-06-09 Fujitsu Ltd センサ補正プログラム、センサ補正装置およびセンサ補正方法
JP2013152165A (ja) * 2012-01-25 2013-08-08 Fujitsu Ltd 検出装置、検出プログラム、及び検出方法
WO2014185027A1 (fr) * 2013-05-15 2014-11-20 旭化成株式会社 Dispositif et procédé d'estimation de décalage et programme
JP2015184160A (ja) * 2014-03-25 2015-10-22 セイコーエプソン株式会社 参照値生成方法、運動解析方法、参照値生成装置及びプログラム

Similar Documents

Publication Publication Date Title
EP3819592B1 (fr) Dispositif de positionnement, procédé de positionnement et programme
KR101851836B1 (ko) 물체의 움직임을 추정하기 위한 시스템 및 방법
Barthold et al. Evaluation of gyroscope-embedded mobile phones
US9599634B2 (en) System and method for calibrating inertial measurement units
Gädeke et al. Smartphone pedestrian navigation by foot-IMU sensor fusion
JP2014506318A (ja) 移動素子または移動体の軌跡推定方法および軌跡推定装置
CN106662443A (zh) 用于垂直轨迹确定的方法和系统
Zhuang et al. A MEMS multi-sensors system for pedestrian navigation
Zhao et al. Adaptive 3D position estimation of pedestrians by wearing one ankle sensor
Montorsi et al. Design and implementation of an inertial navigation system for pedestrians based on a low-cost MEMS IMU
JP5678357B2 (ja) 回転情報演算方法、回転情報演算プログラム、磁気型ジャイロスコープおよび移動体
WO2013057942A1 (fr) Procédé de génération de valeurs de référence et dispositif générateur de valeurs de référence
CN106595669B (zh) 一种旋转体姿态解算方法
Tian et al. Pedestrian navigation system using MEMS sensors for heading drift and altitude error correction
Ayub et al. Sensor placement modes for smartphone based pedestrian dead reckoning
CN109682372B (zh) 一种结合建筑物结构信息与rfid标定的改进型pdr方法
Ali et al. An improved personal dead-reckoning algorithm for dynamically changing smartphone user modes
JP2015017858A (ja) 参照値生成装置及び参照値生成方法
JP2006038650A (ja) 姿勢計測方法、姿勢制御装置、方位計及びコンピュータプログラム
WO2017085756A1 (fr) Dispositif, procédé et programme de traitement d&#39;informations
Nagin et al. Effective integration algorithm for pedestrian dead reckoning
JP5995319B2 (ja) 地磁気を利用して歩行時の鉛直下向きを決定する携帯端末、プログラム及び方法
Yin et al. Pedestrian dead reckoning indoor positioning with step detection based on foot-mounted IMU
CN107289935B (zh) 一种适用于可穿戴设备的室内导航算法
Zhou et al. Pedestrian navigation with foot-mounted inertial sensors in wearable body area networks

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15908696

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15908696

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP