Step detection device

Abstract

To provide a step detection device capable of accurately detecting a step on a road surface.SOLUTION: A step detection device 10 mounted on a wheelchair 100 including a front wheel 101 and a rear wheel 102 has a step detection unit 13 that acquires a pitch rate detected by a gyro sensor 4 for detecting a pitch rate of the wheelchair 100 to detect a step on a road surface on which the wheelchair 100 travels and passes based on the acquired pitch rate. The step detection unit 13 detects a step based on a first pitch rate that is a pitch rate detected by the gyro sensor 4 when the front wheel 101 passes on a road surface and a second pitch rate that is a pitch rate detected by the gyro sensor 4 when the rear wheel 102 passes a position where the first pitch rate is detected.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a step detection device for detecting steps on a road surface.

Background Art

[0002] A barrier-free map has been created that shows steps on the road surface and clearly indicates an easy-to-pass route for users such as wheelchair users. Such barrier-free maps have been surveyed manually for steps and the like, which is time-consuming, costly, and labor-intensive. Therefore, barrier-free maps have only been created for some cities.

[0003] Also, due to manual surveys, it is difficult to frequently update the barrier-free map. Therefore, changes in the road surface conditions such as road construction may not be reflected in the barrier-free map, and there may be a significant difference from the actual road surface conditions.

[0004] In response to such problems, devices that automatically detect steps and the like have been proposed. For example, Patent Document 1 describes that an angular velocity sensor of a mobile terminal detects the angular velocity of a user while walking, compares it with the angular velocity data of a flat place stored in advance, and if the difference is large, determines it as a step of a staircase and transmits the step information and position to a center.

[0005] Also, Patent Document 2 describes that when the road gradient angle of a vehicle calculated by a 3D gyroscope during traveling is larger than a predetermined value, it is determined as a steep step, and the point determined as a step is registered.

[0006] Also, Patent Document 3 describes that a wheelchair is provided with a vibration detection means and an inclination detection means, and the unevenness and inclination of the road surface when moving are uploaded to a server together with position information.

Prior Art Documents

Patent Documents

[0007] [Patent Document 1] Japanese Patent Publication No. 2012-098939 [Patent Document 2] Japanese Patent Publication No. 2005-043261 [Patent Document 3] Japanese Patent Publication No. 2003-010257 [Overview of the project] [Problems that the invention aims to solve]

[0008] The method described in Patent Document 1 can detect steps that are continuous or have a large difference in height, such as stairs, but it may not detect small steps because pedestrians can step over them in one go. The method described in Patent Document 2 distinguishes between steps and slopes based on the magnitude of the gradient angle, making it difficult to accurately detect steps that are relatively low in height and have a small gradient angle. Thus, the methods described in these documents have difficulty accurately detecting steps that do not pose a problem for pedestrians but pose a barrier for wheelchairs and other mobility devices.

[0009] Furthermore, while the method described in Patent Document 3 discloses detecting the unevenness of the road surface using vibration detection means, it does not disclose any specific method for detecting steps or unevenness. Therefore, it is impossible to distinguish whether the detected point is an uneven surface or a step or unevenness.

[0010] Therefore, in view of the above-mentioned problems, the object of the present invention is to provide a step detection device that can accurately detect steps on the road surface. [Means for solving the problem]

[0011] To solve the above problems, the invention described in claim 1 is a step detection device mounted on or attachable to a moving body having front wheels and rear wheels, comprising: acceleration acquisition means for acquiring acceleration detected by acceleration detection means for detecting acceleration perpendicular to the plane of movement of the moving body; and step detection means for acquiring pitch rate detected by pitch rate detection means for detecting pitch rate of the moving body, and detecting a step on the road surface over which the moving body has traveled based on the acquired pitch rate and the acceleration acquired by the acceleration acquisition means, wherein the step detection means detects the step based on a first pitch rate which is the pitch rate detected by the pitch rate detection means when the front wheels pass over the road surface; acceleration acquired in the acceleration acquisition step when the first pitch rate is detected; a second pitch rate which is the pitch rate detected by the pitch rate detection means when the rear wheels pass over the position where the first pitch rate was detected; and acceleration acquired by the acceleration acquisition means when the second pitch rate is detected.

[0012] Furthermore, the invention described in claim 9 is a step detection method performed by a step detection device mounted on or attachable to a moving body having front wheels and rear wheels, the step detection method comprising: an acceleration acquisition step of acquiring acceleration detected by acceleration detection means for detecting acceleration perpendicular to the moving plane of the moving body; and a step detection step of acquiring pitch rate detected by pitch rate detection means for detecting pitch rate of the moving body, and detecting a step on the road surface over which the moving body has traveled based on the acquired pitch rate and the acceleration acquired in the acceleration acquisition step, wherein the step detection step is characterized by detecting the step based on a first pitch rate which is the pitch rate detected by the pitch rate detection means when the front wheels pass over the road surface; the acceleration acquired in the acceleration acquisition step when the first pitch rate is detected; a second pitch rate which is the pitch rate detected by the pitch rate detection means when the rear wheels pass over the position where the first pitch rate was detected; and the acceleration acquired in the acceleration acquisition step when the second pitch rate is detected.

[0013] Furthermore, the invention described in claim 10 is a step detection program characterized by causing a computer to execute the step detection method described in claim 9.

[0014] Furthermore, the invention described in claim 11 is a computer-readable recording medium characterized by storing the step detection program described in claim 10. [Brief explanation of the drawing]

[0015] [Figure 1] This is a schematic diagram of a step detection system having a step detection device according to one embodiment of the present invention. [Figure 2] Figure 1 is a schematic diagram of the configuration of the arithmetic unit 1, etc. [Figure 3] Figure 1 is a schematic diagram of the server device configuration. [Figure 4] This is an explanatory diagram showing how a wheelchair moves when encountering an upward step. [Figure 5] It is a graph showing changes in pitch rate, pitch angle, and vertical acceleration in the state of FIG. 4. [Figure 6] It is an explanatory diagram showing the movement state of a wheelchair in the case of a downward step. [Figure 7] It is a graph showing changes in pitch rate, pitch angle, and vertical acceleration in the state of FIG. 6. [Figure 8] It is an explanatory diagram showing the relationship between the height of a step and the wheelbase length. [Figure 9] It is a graph of the peak values of pitch rate before and after normalizing with the wheelbase length. [Figure 10] It is a table summarizing the average value, standard deviation, and coefficient of variation calculated from the graph shown in FIG. 9. [Figure 11] It is a flowchart showing the operation of the arithmetic unit shown in FIG. 1. [Figure 12] It is a flowchart showing the operation of the server device shown in FIG. 1.

Mode for Carrying Out the Invention

[0016] Hereinafter, a step detection device according to an embodiment of the present invention will be described. The step detection device according to an embodiment of the present invention is mounted on a moving body having front wheels and rear wheels or can be attached to the moving body. And the step detection means is based on a first pitch rate that is the pitch rate detected by the pitch rate detection means when the front wheel passes on the road surface, and a second pitch rate that is the pitch rate detected by the pitch rate detection means when the rear wheel passes the position where the first pitch rate was detected, to detect a step. By doing so, since a step can be detected by the change in the pitch rate of the two wheels of the front wheel and the rear wheel, even a relatively low step can be accurately detected on the road surface.

[0017] Furthermore, the step detection means may use the pitch rate detected when the moving body moves a distance related to the wheelbase length from the position on the road surface where the first pitch rate was detected as the second pitch rate. In this way, the position for detecting the second pitch rate can be determined based on the wheelbase length, which is the length between the front and rear wheels of the moving body and is a known value.

[0018] Furthermore, the step detection means may detect a position as a step if the absolute value of the first pitch rate is greater than or equal to a predetermined first threshold, and the absolute value of the second pitch rate is greater than or equal to a predetermined second threshold. By doing so, a step can be identified when a pitch rate with an absolute value above a certain level is detected when the front wheel passes and when the rear wheel passes, thus enabling accurate step detection.

[0019] Furthermore, the step detection means may include an acceleration acquisition means that acquires the acceleration detected by an acceleration detection means that detects the acceleration perpendicular to the plane of movement of the moving body, and the step detection means may detect the step based on the acceleration acquired by the acceleration acquisition means when the first pitch rate is detected and the acceleration acquired by the acceleration acquisition means when the second pitch rate is detected. By doing so, in addition to the pitch rate, the vertical acceleration can also be taken into consideration, so the accuracy of step detection can be further improved.

[0020] Furthermore, the step detection means may detect a position as a step if the absolute value of the acceleration acquired by the acceleration acquisition means when the first pitch rate is detected is greater than or equal to a predetermined third threshold, and the absolute value of the acceleration acquired by the acceleration acquisition means when the second pitch rate is detected is greater than or equal to a predetermined fourth threshold. In this way, the first pitch rate can be determined when a certain level of vertical acceleration is detected when the front wheel passes and when the rear wheel passes, thereby enabling accurate detection of steps.

[0021] Furthermore, the step detection means may detect the step based on the sign of the pitch rate obtained from the pitch rate detection means and the direction of the change in acceleration obtained from the acceleration acquisition means. By doing so, the sign of the pitch rate, that is, whether the pitch rate is positive or negative, and the direction of the change in acceleration, that is, whether the acceleration increased or decreased, can be taken into consideration, so that the step can be detected with high accuracy by taking into account the direction of the pitch rate and the direction of acceleration.

[0022] Furthermore, the step detection means may detect whether the step over which the moving object has passed was an upward step or a downward step based on the sign of the first pitch rate and the sign of the second pitch rate. In this way, it is possible to determine whether the step currently over which the object has passed is an upward step or a downward step based on the signs of the two pitch rates.

[0023] Furthermore, when a step is detected, the step detection means generates normalized pitch rate information including a first normalized pitch rate obtained by normalizing the peak value of the first pitch rate by the wheelbase length, and / or a second normalized pitch rate obtained by normalizing the peak value of the second pitch rate by the wheelbase length.The position acquisition means may then acquire information about the location where the step was detected, and the transmission means may transmit the information about the location where the step was detected and the normalized pitch rate information obtained from the position acquisition means to an external server device as information for the server device to determine the level of the step on the road surface.By doing so, detection errors due to differences in wheelbase length can be reduced by using the normalized pitch rate information.In addition, the server device can store the results of determining the step level based on information collected from multiple step detection devices.This makes it possible to improve the accuracy of step detection.

[0024] Furthermore, the step detection means may determine the level of the step based on a first normalized pitch rate obtained by normalizing the peak value of the first pitch rate by the wheelbase length, and / or a second normalized pitch rate obtained by normalizing the peak value of the second pitch rate by the wheelbase length, when a step is detected. By doing so, detection errors due to differences in wheelbase length can be reduced.

[0025] Alternatively, the position acquisition means may acquire information regarding the location where a step is detected, and the transmission means may transmit the step level determined by the step detection means and the location where the step is detected to an external server device. In this way, the server device can store the results of determining the step level based on information collected from multiple step detection devices.

[0026] Furthermore, in the step detection method according to one embodiment of the present invention, in the step detection step, a step is detected based on a first pitch rate, which is the pitch rate detected by the pitch rate detection means when the front wheel passes over the road surface, and a second pitch rate, which is the pitch rate detected by the pitch rate detection means when the rear wheel passes over the position where the first pitch rate was detected. In this way, a step can be detected by the change in the pitch rates of the two wheels, the front wheel and the rear wheel, so even relatively low steps on the road surface can be detected with high accuracy.

[0027] Alternatively, the above-described step detection method may be implemented as a step detection program using a computer. By doing so, the computer can detect steps based on changes in the pitch rates of the two wheels, the front and rear wheels, thereby enabling accurate detection of steps on the road surface.

[0028] Furthermore, the aforementioned step detection program may be stored on a computer-readable recording medium. This allows the program to be distributed independently, in addition to being incorporated into equipment, and facilitates version upgrades and other modifications. [Examples]

[0029] An embodiment of the present invention, a step detection device, will be described with reference to Figures 1 to 12. The calculation device 1, which serves as the step detection device in this embodiment, is mounted on a wheelchair 100, which is a mobile device, as shown in Figure 1.

[0030] Figure 1 is a diagram showing the configuration of a step detection system having a step detection device according to one embodiment of the present invention. As shown in Figure 1, the wheelchair 100 is equipped with a GPS receiver 2, an acceleration sensor 3 as an acceleration detection means, a gyro sensor 4 as a pitch rate detection means, and a communication device 5 as a transmission means, in addition to the computing device 1. Here, the computing device 1 and the communication device 5 constitute the step detection device 10 according to one embodiment of the present invention.

[0031] The communication device 5 mounted on the wheelchair 100 can connect wirelessly to a network N such as the Internet, and can communicate with the server device 50 via this network N.

[0032] Wheelchair 100 is equipped with a pair of front wheels 101 and a pair of rear wheels 102 on its body. The front wheels 101 are located on the front side of the wheelchair body. The rear wheels 102 are located on the rear side of the wheelchair body. The diameter of the front wheels 101 is smaller than the diameter of the rear wheels 102.

[0033] The wheelchair 100's frame structure is made up of, for example, a steel pipe frame. The frame is equipped with a seat for the user to sit on, a footplate for the user's feet, and other such features.

[0034] Figure 2 shows a functional configuration diagram of the equipment mounted on wheelchair 100. The computing unit 1 includes, for example, a microcomputer equipped with a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), and interfaces for connecting to a GPS receiver 2, an acceleration sensor 3, a gyro sensor 4, and a communication device 5. Furthermore, by executing a control program stored in ROM, etc., the computing unit 1 functions as a pitch rate acquisition unit 11, an acceleration acquisition unit 12 as an acceleration acquisition means, and a step detection unit 13 as a step detection means.

[0035] The pitch rate acquisition unit 11 acquires the pitch rate detected by the gyro sensor 4. The pitch rate acquisition unit 11 also acquires the pitch rate when it meets the conditions described later, as the first pitch rate and the second pitch rate.

[0036] The acceleration acquisition unit 12 acquires acceleration and other data detected by the acceleration sensor 3.

[0037] The step detection unit 13 detects steps on the road surface based on the pitch rates (first pitch rate and second pitch rate) acquired by the pitch rate acquisition unit 11. The step detection method will be described later. In other words, the step detection unit 13 functions as a step detection means that detects steps on the road surface that the mobile body (wheelchair 100) has traveled over, based on the acquired pitch rates. In this embodiment, a step refers to a position (point) on the road surface where it is difficult for the wheelchair 100 to go up or down, and where one road surface and the next road surface are connected by a vertical surface of a predetermined height or a slope of a predetermined angle or more.

[0038] As is well known, GPS receiver 2 uses multiple GPS (Global Positioning System) satellites or The device receives the transmitted radio waves, calculates the current location information (latitude and longitude), and outputs it to the computing device 1.

[0039] The acceleration sensor 3 detects acceleration in a direction perpendicular to the plane of movement (vertical acceleration) as the wheelchair 100 travels. The acceleration sensor 3 can be any type of sensor, such as a capacitive or piezoresistive type, but it is preferable that it be small since it is mounted on the wheelchair 100.

[0040] The gyro sensor 4 acquires the rotational angular velocity (pitch rate) of the wheelchair 100 in the pitch direction. Here, pitch refers to the tilt in the vertical direction relative to the direction of travel of the wheelchair 100 (rotation angle around the horizontal axis). The gyro sensor 4 can be any type of sensor, such as a capacitive or piezoelectric type, but it is preferable that it be small since it is mounted on the wheelchair 100.

[0041] The communication device 5 transmits the results calculated by the computing device 1 to the server device 50 via wireless communication. The communication device 5 may use communication methods used in mobile phone networks such as LTE (Long Term Evolution) or W-CDMA (Wideband Code Division Multiple Access). It may also use wireless LAN communication methods such as Wi-Fi (registered trademark), or it may be capable of switching between these methods.

[0042] Figure 3 shows a functional configuration diagram of the server device 50. The server device 50 comprises a communication device 51, an arithmetic unit 52, and a storage device 53. The communication device 51 receives the results calculated by the arithmetic unit 1 transmitted from the communication device 5.

[0043] The arithmetic unit 52 has, for example, a microcomputer equipped with a CPU (Central Processing Unit), ROM (Read Only Memory), and RAM (Random Access Memory). The arithmetic unit 52 also functions as a level determination unit 521 and an update unit 522 by executing a control program stored in the ROM, etc.

[0044] The level determination unit 521 determines the level of the step based on the information transmitted from the arithmetic unit 1. The level of the step is a classification of the height of the step into levels, for example, level 1 for 0 to less than 5 cm, level 2 for 5 cm to less than 10 cm, and so on, based on predetermined thresholds.

[0045] The update unit 522 updates the barrier information of the map information 532 stored in the storage device 53 based on the determination result of the level determination unit 521. The update unit 522 also stores information such as the normalized pitch rate transmitted from the arithmetic unit 1 as normalized pitch rate information 531.

[0046] Next, the principle of step detection in this embodiment will be explained with reference to Figures 4 to 7. Figure 4 shows the movement state of wheelchair 100 when passing over an upward step U, and Figure 5 is a graph showing the changes in pitch rate, pitch angle, and vertical acceleration in the state shown in Figure 4. Note that the vertical acceleration graph in Figure 5(c) has an initial value of approximately 9.8 m / s², and the initial value is the acceleration due to gravity. In other words, Figure 5(c) means that downward acceleration has been detected in advance, and as the value increases, the downward acceleration increases.

[0047] First, the state shown in Figure 4(a) is before passing over the upward step U. In this case, there are no significant changes in pitch rate, pitch angle, or vertical acceleration (Figure 5(a)a). Next, the state shown in Figure 4(b) is when the front wheels 101 are going up the upward step U (the front wheels 101 have collided with the step). In this case, as the front wheels 101 go up the upward step U, the body of the wheelchair 100 is tilted diagonally upward in the direction of travel, and the pitch rate increases upward (Figure 5(a)b). Also, an upward force is received from the upward step U, causing an upward movement, which generates an upward vertical acceleration, and the vertical acceleration increases (Figure 5(c)b).

[0048] Next, the state shown in Figure 4(c) is when the rear wheels 102 are going up the step U (the rear wheels 102 have collided with the step). In this case, as the rear wheels 102 go up the step U, the body of the wheelchair 100 tries to transition from an inclined state to a horizontal state, and the pitch rate increases downward (Figure 5(a)c). Also, as the rear wheels 102 go up the step, the body receives an upward force from the step U, so the vertical acceleration increases (Figure 5(c)c). Finally, the state shown in Figure 4(d) is after passing the step U. In this case as well, similar to before passing the step U, there are no significant changes in pitch rate, pitch angle, or vertical acceleration (Figure 5(a)d).

[0049] In other words, in the case of an upward step U, when the front wheels 101 pass over the upward step U, the pitch rate increases in the positive direction and the vertical acceleration increases. That is, the vertical acceleration obtained by subtracting the acceleration due to gravity from the pitch rate is a positive number. Then, when the rear wheels 102 pass over the upward step U, the pitch rate increases in the negative direction and the vertical acceleration increases. That is, the pitch rate becomes a negative number, and the vertical acceleration obtained by subtracting the acceleration due to gravity is a positive number. Here, in this embodiment, the pitch rate is a positive number when the wheelchair 100 is tilted upward relative to the direction of travel, and a negative number when the wheelchair is tilted downward relative to the direction of travel. Also, the vertical acceleration obtained by subtracting the acceleration due to gravity is a positive number when tilting upward and a negative number when tilting downward. Here, as the front wheel 101 and rear wheel 101 pass over the incline U, the vertical acceleration increases and then decreases. This is because the wheels rise to a position higher than the top of the step due to the collision with the step, and then fall from that position to the top of the step (immediately after points b and c in Figure 5(c)).

[0050] Figure 6 shows the movement of wheelchair 100 when passing over a downward step D, and Figure 7 is a graph showing the changes in pitch rate, pitch angle, and vertical acceleration in the state shown in Figure 6.

[0051] First, the state shown in Figure 6(a) is before passing over the downward step D. In this case, there are no significant changes in pitch rate, pitch angle, or vertical acceleration (Figure 7(a)a). Next, the state shown in Figure 6(b) is when the front wheels 101 are descending (falling) over the downward step D. In this case, as the front wheels 101 descend over the downward step D, the body of the wheelchair 100 is angled downwards in the direction of travel, and the pitch rate increases downwards (Figure 7(a)b). Also, as the front wheels 101 fall, the vertical acceleration decreases (Figure 7(c)b).

[0052] Next, Figure 6(c) shows the state where the rear wheels 102 are descending (falling) down the step D. In this case, as the rear wheels 102 descend the step D, the body of the wheelchair 100 attempts to transition from an inclined state to a horizontal state, causing the pitch rate to increase upward (Figure 7(a)c). Also, as the rear wheels 102 fall, the vertical acceleration decreases (Figure 7(c)c). Finally, Figure 6(d) shows the state after passing the step D. In this case as well, similar to before passing the step D, there are no significant changes in pitch rate, pitch angle, or vertical acceleration (Figure 7(a)d).

[0053] In other words, in the case of a downward step D, when the front wheel 101 passes over the step D, the pitch rate increases in the negative direction and the vertical acceleration decreases. That is, the vertical acceleration obtained by subtracting the gravitational acceleration from the pitch rate becomes a negative number. Then, when the rear wheel 102 passes over the downward step D, the pitch rate increases in the positive direction and the vertical acceleration decreases. That is, the pitch rate becomes a positive number and the vertical acceleration obtained by subtracting the gravitational acceleration becomes a negative number. Here, when the front wheel 101 and the rear wheel 102 pass over the downward step D, the vertical acceleration decreases and then begins to increase, which is because of the rebound that occurs after the wheel hits the ground (immediately after points b and c in Figure 7(c)).

[0054] Therefore, as shown in Figures 4 to 7, it is possible to detect whether an upward step U or a downward step D has been passed based on the value and sign of the pitch rate detected by the gyro sensor 4 and the absolute value and direction of change (increase or decrease) of the vertical acceleration detected by the acceleration sensor 3. Furthermore, by setting thresholds for the detected pitch rate and vertical acceleration, it becomes possible to reliably detect the peaks that appear when passing over steps such as points b and c shown in Figures 5 and 7.

[0055] Next, we will explain a method for accurately determining the height level of the step detected using the principle described above. As mentioned above, the presence or absence of a step can be detected by the pitch rate, etc., but the height of the step cannot be determined simply by the peak value of the pitch rate. This is because wheelchair 100 has different wheelbase lengths (the distance between the axle of the front wheel 101 and the axle of the rear wheel 102).

[0056] As shown in Figure 8, if the step height is h, the wheelbase length is H, and the pitch angle is θ, then h = Hsinθ. Also, when the pitch angle θ is small, sinθ ≈ θ, so h = Hθ. Therefore, if h is constant, as the wheelbase length H decreases, the pitch angle θ increases. The same applies to the pitch rate, which is the time change of the pitch angle, so it can be said that there is a correlation (inverse proportion) between the wheelbase H and the peak value of the pitch rate.

[0057] Therefore, in this embodiment, the influence of the wheelbase length is eliminated by normalizing the detected peak value of the pitch rate by the wheelbase length. One example of normalization is multiplying the peak value of the pitch rate by the reciprocal of the wheelbase length. Note that other normalization methods may also be used.

[0058] Figure 9 shows a comparison of pitch rates before and after normalization. Figure 9(a) shows the case when the step height is 2 cm, and Figure 9(b) shows the case when the step height is 4 cm. The horizontal axis in Figure 9 is the sample number used to identify the wheelchairs on which the data was measured. Also, in Figure 9, the diamonds represent the data before normalization, and the circles represent the data after normalization.

[0059] Figure 10 is a table summarizing the graphs shown in Figure 9. As shown in Figure 10, in the case of a 2cm step, the normalized mean μ decreases from 28.4 to 12.8, the standard deviation σ decreases from 6.98 to 2.50, and the coefficient of variation cv decreases from 0.246 to 0.195. Similarly, in the case of a 4cm step, the normalized mean μ decreases from 44.7 to 19.8, the standard deviation σ decreases from 10.0 to 3.06, and the coefficient of variation cv decreases from 0.225 to 0.155. Here, the coefficient of variation is the standard deviation divided by the mean, and this value can be used to evaluate the degree of variability. A smaller coefficient of variation indicates less variability, while a larger coefficient of variation indicates more variability.

[0060] Therefore, normalization can reduce the variability caused by differences in the wheelbase of wheelchair 100. Thus, the accuracy (reliability) of step level discrimination can be improved.

[0061] Next, the operation of the arithmetic unit 1 and server unit 50 with the above configuration will be explained with reference to the flowcharts in Figures 11 and 12. Figure 11 is a flowchart of the operation of the arithmetic unit 1.

[0062] First, in step S101, thresholds T1, T2, T3, and T4 are set in the step detection unit 13, and the process proceeds to step S102. Thresholds T3 and T4 are set to the vertical acceleration detected by the acceleration sensor 3. Threshold T3 is the threshold for detecting when the front wheels 101 have passed over a step, and threshold T4 is the threshold for detecting when the rear wheels 102 have passed over a step. Thresholds T1 and T2 are set to the pitch rate detected by the gyro sensor 4. Threshold T1 is the threshold for detecting when the front wheels 101 have passed over a step, and threshold T2 is the threshold for detecting when the rear wheels 102 have passed over a step. These thresholds T1, T2, T3, and T4 are adjustable to match the wheelchair 100. Alternatively, a data table corresponding to the wheelchair type may be prepared in advance, and the thresholds T1, T2, T3, T4 and the wheelbase length H may be automatically set by inputting the wheelchair type.

[0063] Next, in step S102, the vertical acceleration Az and pitch rate Pr are detected, and the process proceeds to step S103. In this step, the acceleration acquisition unit 12 acquires the vertical acceleration Az detected by the acceleration sensor 3, and the pitch rate acquisition unit 11 acquires the pitch rate Pr detected by the gyro sensor 4. That is, the pitch rate Pr acquired in this step becomes the first pitch rate.

[0064] Next, in step S103, the step detection unit 13 determines whether the vertical acceleration Az obtained in step S102 is less than -T3 and the pitch rate Pr is less than -T1. If these conditions are met (Yes), the process proceeds to step S104; otherwise, it proceeds to step S105. This step detects point b in Figures 7(a) and 7(c). Note that the vertical acceleration Az compared with the threshold in this step and beyond is the value after subtracting the gravitational acceleration, and the threshold is also defined by the change from the gravitational acceleration. That is, threshold T3 is the third threshold, and threshold T1 is the first threshold. The unit also detects that the signs of the vertical acceleration Az and pitch rate Pr are negative (negative numbers).

[0065] Next, in step S104, since it was determined in step S103 that the conditions were met, the step detection unit 13 tentatively determines that the step currently being passed is a downward step, sets the step candidate flag to -1, and proceeds to step S107. In this step, a state corresponding to point b in Figure 7(a)(c) was detected, but the determination on the rear wheel 102 side has not been made, so a tentative determination is made, and the step candidate flag set in the calculation unit 1 is set as a flag indicating that determination. Note that the value set in the step candidate flag is not limited to -1, but can be any value that indicates that it has been tentatively determined to be a downward step.

[0066] On the other hand, in step S105, the step detection unit 13 determines whether the vertical acceleration Az obtained in step S102 is T3 or greater and the pitch rate Pr is T1 or greater. If these conditions are met (Yes), the process proceeds to step S106. If the conditions are not met (No), the process determines that no step has been detected and terminates this flowchart. This step detects point b in Figures 5(a) and 5(c). That is, it also detects that the signs of the vertical acceleration Az and the pitch rate Pr are positive (positive numbers).

[0067] Next, in step S106, since it was determined in step S105 that the conditions were met, the step detection unit 13 tentatively determines that the step currently being passed is an uphill step, sets the step candidate flag to 1, and proceeds to step S107. In this step, a state corresponding to point b in Figure 5(a)(c) was detected, but the determination on the rear wheel 102 side has not been made, so a tentative determination is made and the step candidate flag is set in the same way as in step S104. Note that the value set in the step candidate flag is not limited to 1, but can be any value that indicates that it has been tentatively determined to be an uphill step.

[0068] Next, in step S107, the vertical acceleration Az and pitch rate Pr after the wheelbase length H has been moved are detected, and the process proceeds to step S108. In this step, similar to step S102, the acceleration acquisition unit 12 acquires the vertical acceleration Az detected by the acceleration sensor 3, and the pitch rate acquisition unit 11 acquires the pitch rate Pr detected by the gyro sensor 4. The distance traveled by the wheelbase length H may be calculated from the latitude and longitude detected by the GPS receiver 2 based on a preset wheelbase length H, or it may be calculated by multiplying the rotation angle of the wheelchair 100's wheels (rear wheels 102) by a preset circumference of the wheels. The rotation angle of the wheelchair 100's wheels can be detected by providing angle sensors or the like on the wheels. In this step, the vertical acceleration Az and pitch rate Pr are detected within a predetermined range (predetermined range) before and after the distance of the wheelbase length H. That is, this predetermined range (predetermined range) before and after the distance of the wheelbase length H is the distance related to the wheelbase length of the moving object (wheelchair 100). The pitch rate Pr obtained in step S108 becomes the second pitch rate.

[0069] Next, in step S108, the step detection unit 13 determines whether the vertical acceleration Az obtained in step S107 is less than -T4 and the pitch rate Pr is T2 or greater. If this condition is met (Yes), the process proceeds to step S109; otherwise, the process proceeds to step S111. This step detects point c in Figures 7(a) and 7(c). That is, threshold T4 becomes the fourth threshold, and threshold T2 becomes the second threshold. The unit also detects that the sign of the vertical acceleration Az is negative (a negative number) and the sign of the pitch rate Pr is positive (a positive number).

[0070] Next, in step S109, the step detection unit 13 determines whether the step candidate flag is -1 or not. If it is -1 (Yes), it proceeds to step S110; otherwise (No), it determines that no step has been detected and terminates this flowchart.

[0071] Next, in step S110, the step candidate flag indicated a downward step, thus satisfying the conditions for points b and c in Figure 7. The step detection unit 13 officially determined that the traversed step was a downward step and proceeded to step S114. Specifically, the absolute value of the first pitch rate is greater than or equal to the first threshold (T1), and the absolute value of the second pitch rate is greater than or equal to the second threshold (T2). Also, the absolute value of the vertical acceleration obtained when the first pitch rate was detected is greater than or equal to the third threshold (T3), and the absolute value of the vertical acceleration obtained when the second pitch rate was detected is greater than or equal to the fourth threshold (T4). The step is then detected based on these conditions. Furthermore, since the first pitch rate is a negative number and the second pitch rate is a positive number, the traversed step is a downward step. In other words, the downward step is detected based on the sign of the first pitch rate and the sign of the second pitch rate.

[0072] On the other hand, in step S111, the step detection unit 13 determines whether the vertical acceleration Az obtained in step S107 is T4 or greater and the pitch rate Pr is less than -T2. If these conditions are met (Yes), the process proceeds to step S112. If the conditions are not met (No), the process determines that no step has been detected and terminates this flowchart. This step detects point c in Figures 5(a) and 5(c). That is, it also detects that the sign of the vertical acceleration Az is positive (positive number) and the sign of the pitch rate Pr is negative (negative number).

[0073] Next, in step S112, the step detection unit 13 determines whether the step candidate flag is 1 or not. If it is 1 (Yes), the process proceeds to step S113. Otherwise (No), it determines that no step has been detected and terminates this flowchart.

[0074] Next, in step S113, the step candidate flag indicated an upward step, thus satisfying the conditions for points a and c in Figure 5. The step determination unit 13 formally determined that the traversed step was an upward step and proceeded to step S114. Specifically, the absolute value of the first pitch rate is greater than or equal to the first threshold (T1), and the absolute value of the second pitch rate is greater than or equal to the second threshold (T2). Also, the absolute value of the vertical acceleration obtained when the first pitch rate was detected is greater than or equal to the third threshold (T3), and the absolute value of the vertical acceleration obtained when the second pitch rate was detected is greater than or equal to the fourth threshold (T4). The step is then detected based on these conditions. Furthermore, since the first pitch rate is a positive number and the second pitch rate is a negative number, the traversed step is an upward step. In other words, the upward step is detected based on the sign of the first pitch rate and the sign of the second pitch rate.

[0075] Next, in step S114, the step detection unit 13 acquires the position (latitude and longitude) from the GPS receiver 2 and records the position of the step in its internal memory or the like. If the wheelbase length was detected using the GPS receiver 2 in step S107, the position acquired at that time may also be recorded. In other words, the step detection unit 13 functions as a position acquisition means that acquires information regarding the position where a step has been detected.

[0076] Next, in step S115, the step detection unit 13 detects the peak value of the pitch rate Pr at the location (position) where it determined there was a step, and proceeds to step S116. The peak value of the pitch rate Pr may be stored in internal memory or the like and used from the values ​​used in the determinations in steps S103, S105, S108, S111, etc. Note that the peak value detected in this step may be the peak value on the front wheel 101 side (when the determination is made in steps S103 and S105), the peak value on the rear wheel 102 side (when the determination is made in steps S108 and S111), or the average of both.

[0077] Next, in step S116, the step detection unit 13 normalizes the value by multiplying the reciprocal of the wheelbase length H of the wheelchair 100 by the peak value of the pitch rate Pr detected in step S115, and proceeds to step S117. The normalized peak value is defined as the normalized pitch rate Pr_n. That is, the normalized pitch rate Pr_n becomes the normalized pitch rate information. Here, the normalized pitch rate Pr_n is the first normalized pitch rate if the peak value of the front wheel 101 side is used in step S115, the second normalized pitch rate if the rear wheel 102 side is used, and the average normalized pitch rate if the average of the peak values ​​of the front wheel 101 side and the rear wheel 102 side is used.

[0078] Next, in step S117, the step detection unit 13 adds the ID of the wheelchair 100 and transmits the position where the step was determined (step position Sp) and the normalized pitch rate value Pr_n to the server device 50 via the communication device 5. The ID of the wheelchair 100 is an ID that has been assigned to each wheelchair in advance and is set in the calculation device 1.

[0079] As is clear from the above explanation, the flowchart in Figure 11 functions as a step detection process that detects a step based on a first pitch rate, which is the pitch rate detected by the pitch rate detection means when the front wheel 101 passes over the road surface, and a second pitch rate, which is the pitch rate detected by the pitch rate detection means when the rear wheel 102 passes over the position where the first pitch rate was detected.

[0080] Next, the operation of the server device 50 shown in Figure 12 will be explained. The flowchart shown in Figure 12 is executed by the arithmetic unit 52.

[0081] First, in step S201, the arithmetic unit 52 reads the ID of wheelchair 100 received by the communication device 51 and obtains the weighted value W of wheelchair 100. Then, the arithmetic unit 52 uses W to average the normalized pitch rate value Pr_n at the step position Sp received by the communication device 51 with past data (taking W into consideration) and proceeds to step S202. The value calculated by this averaging process is called the average value Pr_ave.

[0082] This weighted value W indicates the confidence level of previously transmitted normalized pitch rate values ​​Pr_n and is stored in the memory device 53, linked to the ID of wheelchair 100. This weighted value W can be set to, for example, 1 initially and increased as the confidence level increases. Then, a weighted average is calculated together with the normalized pitch rate information 531 data stored in the memory device 53, which contains past data.

[0083] Next, in step S202, the arithmetic unit 52 changes the weighting value W for the wheelchair 100 of the ID received in step S201 and proceeds to step S203. In this step, if the difference between the average value Pr_ave and the normalized pitch rate value Pr_n received by the communicator 51 is large, the weighting value W for this wheelchair 100 is decreased by a predetermined amount. Conversely, if the difference between the average value Pr_ave and the normalized pitch rate value Pr_n received by the communicator 51 is small, the weighting value W for this wheelchair 100 is increased by a predetermined amount. In other words, wheelchairs that output a normalized pitch rate value Pr_n close to the average value Pr_ave will have a higher weighting value W.

[0084] Next, in step S203, the level determination unit 521 of the calculation unit 52 compares the threshold value of the step level with the average value Pr_ave calculated in step S202, determines the step level, and proceeds to step S204. This threshold value is set in advance according to the step level, for example, to 5 cm, 10 cm, etc. For example, if the calculated average value Pr_ave is 6.8 cm, it falls within the range of 5 cm or more and less than 10 cm, so it is determined to be step level 2, etc.

[0085] Next, in step S204, the update unit 522 of the arithmetic unit 52 creates or updates barrier information stored in the map information 532 of the storage device 53 (creates or updates a barrier-free map) based on the result of the step level determined in step S203.

[0086] As explained above, determining the step level requires the peak value of the pitch rate. However, if you simply want to determine the step level, the peak value is not necessary; the determination can be made when a value equal to or greater than the respective thresholds T1, T2, T3, and T4 is detected.

[0087] Furthermore, in the flowchart described above, the thresholds for determining whether a step is an upward or downward step were the same absolute value with different polarity (e.g., T1 and -T1), but different thresholds may be set for each. As previously mentioned, when the front wheel 101 and rear wheel 101 pass over an upward step U, the vertical acceleration increases and then decreases because the wheels rise to a position higher than the top of the step due to collision with the step, and then fall from that position to the top of the step. When the front wheel 101 and rear wheel 101 pass over a downward step D, the vertical acceleration decreases and then increases because the wheels bounce back after colliding with the ground. Therefore, by setting different thresholds for these changes in vertical acceleration and making determinations accordingly, it is possible to detect upward and downward steps with even greater accuracy.

[0088] In this embodiment, the computing unit 1 is mounted on a wheelchair 100 equipped with front wheels 101 and rear wheels 102. The pitch rate detected by the gyro sensor 4 when the front wheels 101 pass over the road surface and acquired by the pitch rate acquisition unit 11 is designated as the first pitch rate. The pitch rate detected by the gyro sensor 4 when the rear wheels 102 pass over the position where the first pitch rate was detected and acquired by the pitch rate acquisition unit 11 is designated as the second pitch rate. The step detection unit 13 then detects the step based on the first and second pitch rates. In this way, the step can be detected by the change in the pitch rates of the two wheels, the front wheels 101 and the rear wheels 102, so even relatively low steps on the road surface can be detected with high accuracy.

[0089] Furthermore, the step detection unit 13 uses the pitch rate acquired by the pitch rate acquisition unit 11 as the second pitch rate when the wheelchair 100 moves within a predetermined range in front of and behind a position that has moved a distance including a predetermined range in front of and behind the wheelbase length H of the wheelchair 100 from the position on the road surface where the first pitch rate was detected. In this way, the position for detecting the second pitch rate can be identified based on the wheelbase length H, which is the length between the front wheels 101 and the rear wheels 102 of the wheelchair 100 and is a known value.

[0090] Furthermore, the step detection unit 13 detects a position as a step when the absolute value of the first pitch rate is greater than or equal to the threshold T1, and the absolute value of the second pitch rate is greater than or equal to the threshold T3. In this way, steps can be identified when a pitch rate with an absolute value greater than a certain level is detected when the front wheel 101 passes and when the rear wheel 102 passes, thus enabling accurate step detection.

[0091] Furthermore, the system includes an acceleration acquisition unit 12 that acquires the vertical acceleration detected by an acceleration sensor 3 that detects acceleration perpendicular to the plane of movement of the wheelchair 100. The step detection unit 13 detects the step based on the vertical acceleration acquired by the acceleration acquisition unit 12 when the first pitch rate is detected and the vertical acceleration acquired by the acceleration acquisition unit 12 when the second pitch rate is detected. In this way, in addition to the pitch rate, vertical acceleration can also be considered, which further improves the accuracy of step detection.

[0092] Furthermore, the step detection unit 13 detects a position as a step if the absolute value of the vertical acceleration acquired by the acceleration acquisition unit 12 when the first pitch rate is detected is greater than or equal to the threshold T3, and the absolute value of the vertical acceleration acquired by the acceleration acquisition unit 12 when the second pitch rate is detected is greater than or equal to the threshold T4. In this way, the first pitch rate can be determined when a certain level of vertical acceleration is detected when the front wheel 101 passes and when the rear wheel 102 passes, thereby enabling accurate step detection.

[0093] Furthermore, the step detection unit 13 detects the step based on the sign of the pitch rate obtained from the gyro sensor 4 and the direction of change in vertical acceleration obtained from the acceleration acquisition unit 12. In this way, the sign of the pitch rate, that is, whether the pitch rate is positive or negative, and the sign of the acceleration, that is, whether the acceleration has increased or decreased, can be taken into consideration, so that the step can be detected with high accuracy by taking into account the direction of the pitch rate and the direction of acceleration.

[0094] Furthermore, the step detection unit 13 detects whether the step the wheelchair 100 has passed over is an upward or downward step based on the sign of the first pitch rate and the sign of the second pitch rate. In this way, it is possible to determine whether the step currently passed over is an upward or downward step based on the signs of the two pitch rates.

[0095] Furthermore, when the step detection unit 13 detects a step, it generates a normalized pitch rate Pr_n by normalizing the peak value of the pitch rate Pr by the wheelbase length H. The step detection unit 13 then obtains the latitude and longitude of the location where the step was detected from the GPS receiver 2. The communication device 5 then transmits the normalized pitch rate Pr_n and the ID of the wheelchair 100 to the server device 50 as information for the server device 50 to determine the level of the step on the road surface, in addition to the latitude and longitude of the location where the step was detected. By doing this, the detection error due to differences in wheelbase length H can be reduced by calculating the normalized pitch rate Pr_n. In addition, the server device 50 can store the results of the step level determination based on information collected from multiple step detection devices. Therefore, the accuracy of step determination can be improved.

[0096] Furthermore, the arithmetic unit 52 of the server device 50 weights the normalized pitch rate Pr_n transmitted from the wheelchair 100 (arithmetic unit 1) based on the weighting value W before determining the step level. By doing so, fluctuations in the step level due to unreliable data are reduced, and the reliability of the step level can be improved.

[0097] In the above-described embodiment, the server device 50 performed the step level determination, but the calculation device 1 mounted on the wheelchair 100 may perform the determination based on the normalized pitch rate Pr_n. The determination result, the latitude and longitude of the location where the step was detected, and the ID of the wheelchair 100 may then be transmitted to the server device 50 via the communication device 5.

[0098] Furthermore, although the above-described embodiment explained the step detection device as a computing device 1 mounted on a wheelchair 100, it does not necessarily have to be a dedicated device. For example, any terminal device with communication capabilities, such as a smartphone equipped with (or capable of connecting to) a GPS receiver, gyro sensor, and accelerometer, can function as a step detection device by making the above-described flowchart into an application (computer program). In this case, a holder or the like can be attached to the wheelchair, and the smartphone or the like can be attached to that holder. In other words, it becomes a step detection device that can be attached to a mobile object.

[0099] Furthermore, in the above-described embodiment, the step was detected based on the detection results of both the acceleration sensor 3 and the gyro sensor 4. However, it is possible to detect the step using only the detection result of the gyro sensor 4, although the accuracy will be lower. In this case, only thresholds T1 and T2 need to be set.

[0100] Furthermore, in the above-described embodiment, information can be obtained as to whether an upward or downward step has been passed. Therefore, the type of step (whether an upward or downward step occurs from which direction of movement) can be determined based on this information and the direction of movement, such as the movement trajectory of the wheelchair 100.

[0101] Furthermore, although the above-described embodiment used a wheelchair as the mobile device, any device equipped with front and rear wheels for travel on a road surface would suffice, such as a mobility scooter, stroller, golf cart, bicycle, automobile, trolley, or robot with wheels.

[0102] Furthermore, the present invention is not limited to the above embodiments. That is, those skilled in the art can implement the invention in various modifications without departing from the core principles, in accordance with conventionally known knowledge. As long as such modifications still possess the configuration of the step detection device of the present invention, they are of course included within the scope of the present invention. [Explanation of Symbols]

[0103] 1 Computing device 2 PS receivers 3. Acceleration sensor (means for acquiring acceleration) 4. Gyroscope sensor 5. Communication device (transmission means) 10. Step detection device 11 Pitch rate acquisition unit 12 Acceleration acquisition unit (acceleration acquisition means) 13. Step detection unit (step detection means, position acquisition means) 50 Server Devices 100 Wheelchairs (mobility devices) 101 Front Wheel 102 Rear wheel

Claims

1. A step detection device mounted on or attachable to a mobile body having front wheels and rear wheels, An acceleration acquisition means that acquires the acceleration detected by an acceleration detection means that detects the acceleration perpendicular to the plane of movement of the moving body, A pitch rate detection means for detecting the pitch rate of the moving body acquires the pitch rate detected by the pitch rate detection means, and a step detection means for detecting a step on the road surface that the moving body has traveled over, based on the acquired pitch rate and the acceleration acquired by the acceleration acquisition means. Equipped with, The step detection means detects the step based on a first pitch rate, which is the pitch rate detected by the pitch rate detection means when the front wheel passes over the road surface; the acceleration acquired by the acceleration acquisition means when the first pitch rate is detected; a second pitch rate, which is the pitch rate detected by the pitch rate detection means when the rear wheel passes over the position where the first pitch rate was detected; and the acceleration acquired by the acceleration acquisition means when the second pitch rate is detected. A step detection device characterized by the following features.

2. The step detection device according to claim 1, characterized in that the step detection means detects a position as a step when the absolute value of the first pitch rate is greater than or equal to a predetermined first threshold, the absolute value of the second pitch rate is greater than or equal to a predetermined second threshold, the absolute value of the acceleration acquired by the acceleration acquisition means when the first pitch rate is detected is greater than or equal to a predetermined third threshold, and the absolute value of the acceleration acquired by the acceleration acquisition means when the second pitch rate is detected is greater than or equal to a predetermined fourth threshold.

3. The step detection device according to claim 1 or 2, characterized in that the step detection means detects the step based on the sign of the pitch rate obtained from the pitch rate detection means and the direction of change in acceleration obtained from the acceleration acquisition means.

4. The step detection device according to any one of claims 1 to 3, characterized in that the step detection means detects whether the step over which the moving body passed was an upward step or a downward step based on the sign of the first pitch rate and the sign of the second pitch rate.

5. The step detection device according to any one of claims 1 to 4, characterized in that the step detection means sets the pitch rate detected when the moving body moves a distance related to the wheelbase length of the moving body from the position on the road surface where the first pitch rate was detected as the second pitch rate.

6. When the step detection means detects the step, it generates normalized pitch rate information based on a first normalized pitch rate obtained by normalizing the peak value of the first pitch rate by the wheelbase length, and / or a second normalized pitch rate obtained by normalizing the peak value of the second pitch rate by the wheelbase length. A position acquisition means for acquiring information regarding the location where the step difference is detected, A transmission means transmits to an external server device information regarding the location where the step was detected and normalized pitch rate information acquired by the position acquisition means, as information for an external server device to determine the level of the step on the road surface. The step detection device according to claim 5, further comprising the following:

7. The step detection device according to claim 5, characterized in that when the step detection means detects the step, it determines the level of the step based on a first normalized pitch rate obtained by normalizing the peak value of the first pitch rate by the wheelbase length, and / or a second normalized pitch rate obtained by normalizing the peak value of the second pitch rate by the wheelbase length.

8. A position acquisition means for acquiring information regarding the location where the step difference is detected, A transmission means for transmitting information regarding the level of the step and the location where the step was detected, as determined by the step detection means, to an external server device. The step detection device according to claim 7, characterized by comprising the following:

9. A step detection method performed by a step detection device mounted on or attachable to a moving body having front wheels and rear wheels, An acceleration acquisition step is to acquire the acceleration detected by acceleration detection means that detects the acceleration perpendicular to the plane of movement of the moving body, A step detection step is performed to obtain the pitch rate detected by a pitch rate detection means for detecting the pitch rate of the moving body, and to detect the step on the road surface that the moving body has traveled over based on the obtained pitch rate and the acceleration obtained in the acceleration acquisition step. Includes, The step detection step detects the step based on a first pitch rate, which is the pitch rate detected by the pitch rate detection means when the front wheel passes over the road surface; the acceleration acquired in the acceleration acquisition step when the first pitch rate is detected; a second pitch rate, which is the pitch rate detected by the pitch rate detection means when the rear wheel passes over the position where the first pitch rate was detected; and the acceleration acquired in the acceleration acquisition step when the second pitch rate is detected. A step detection method characterized by the above.

10. A step detection program characterized by causing a computer to execute the step detection method described in claim 9.

11. A computer-readable recording medium characterized by storing the step detection program described in claim 10.

Images