JP2012154769A - Acceleration detection method, position calculation method and acceleration detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a new method for more appropriately detecting acceleration of a moving body.SOLUTION: An attitude adjustment section 20 adjusts a detection attitude detected by an attitude sensor 4 installed on a moving body using a detection acceleration vector detected by an acceleration sensor 3 installed on the moving body as well as a given reference vector relevant to a movement of the moving body. A coordinate conversion section 30 obtains a conversion coefficient related to coordinate conversion from a local coordinate system of the moving body to an absolute coordinate system using a detection attitude adjusted by the attitude adjustment section 20 and converts the detection acceleration vector detected by the acceleration sensor 3 in the local coordinate system into the acceleration vector in the absolute coordinate system.

Description

  The present invention relates to an acceleration detection method, a position calculation method, and an acceleration detection device.

  In various fields such as so-called seamless positioning, motion sensing, and attitude control, the use of inertial sensors is attracting attention. As an inertial sensor, an acceleration sensor, a gyro sensor, a pressure sensor, a geomagnetic sensor, and the like are widely known. There has also been devised a technique for calculating the position by performing inertial navigation calculation using the detection result of the inertial sensor.

  In inertial navigation calculation, there is a problem that the accuracy of position calculation is reduced due to various error components that can be included in the detection result of the inertial sensor, and there are various techniques for improving the accuracy of position calculation. It has been devised. For example, Patent Document 1 discloses a technique for correcting data measured by performing an inertial navigation calculation using a GPS (Global Positioning System).

JP 2004-239634 A

  In the technique disclosed in Patent Document 1, a Kalman filter is used for each of various quantities related to the inertial navigation calculation such as the position, velocity, acceleration, posture angle, and azimuth of the moving body measured by performing the inertial navigation calculation. Thus, the error that can be included in each quantity is estimated. In this case, since it is necessary to perform error estimation calculation using the Kalman filter for all the quantities related to the inertial navigation calculation, the calculation amount tends to be enormous.

  In inertial navigation calculation, for example, if an acceleration sensor installed on a moving body is used, the speed of the moving body is calculated by integrating the detected acceleration detected by the acceleration sensor, and the position is calculated by integrating the speed. Is calculated.

  In this case, the error that can be included in the calculated speed and position originates from an error that can be included in the acceleration detected by the acceleration sensor. That is, in order to repeat the integration with respect to the detected acceleration, the error included in the detected acceleration is superimposed and accumulated as an error in the speed and position as time elapses. Therefore, if the acceleration of the moving body can be detected correctly, it is possible to improve the accuracy of the inertial navigation calculation without estimating all the errors related to the inertial navigation calculation as in Patent Document 1. it is conceivable that.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to propose a new method for more appropriately detecting the acceleration of a moving body.

  In a first embodiment for solving the above-described problems, a detected attitude of an attitude sensor installed on a moving body is determined based on a detected acceleration vector of an acceleration sensor installed on the moving body and the movement of the moving body. Adjusting using the reference vector, and using the adjusted detection posture, obtaining a conversion coefficient related to coordinate conversion from a local coordinate system associated with the acceleration sensor to an absolute coordinate system; And converting the detected acceleration vector into an acceleration vector in the absolute coordinate system using the conversion coefficient.

  As another form, the detected posture of the posture sensor installed on the moving body is detected using the detected acceleration vector of the acceleration sensor installed on the moving body and a given reference vector related to the movement of the moving body. An adjustment unit for adjusting, a conversion coefficient calculation unit for obtaining a conversion coefficient related to coordinate conversion from a local coordinate system to an absolute coordinate system associated with the acceleration sensor, using the adjusted detection posture; and the conversion You may comprise the acceleration detection apparatus provided with the conversion part which carries out coordinate conversion of the said detected acceleration vector into the acceleration vector in the said absolute coordinate system using a coefficient.

  According to the first aspect and the like, the detected attitude of the attitude sensor installed on the moving body is determined using the detected acceleration vector of the acceleration sensor installed on the moving body and the given reference vector related to the movement of the moving body. Adjust. Then, using the adjusted detection posture, a conversion coefficient related to coordinate conversion from the local coordinate system associated with the acceleration sensor to the absolute coordinate system is obtained. Then, using the obtained conversion coefficient, the detected acceleration vector of the acceleration sensor is coordinate-converted to an acceleration vector in the absolute coordinate system.

  The acceleration vector detected by the acceleration sensor is an acceleration vector in a local coordinate system associated with the acceleration sensor. On the other hand, for example, in inertial navigation calculation, the calculation is generally performed based on an absolute coordinate system that defines the moving space of the moving body. Since an error may be included in the detected posture of the posture sensor, the detected posture of the posture sensor is adjusted using the detected acceleration vector of the acceleration sensor and a given reference vector related to the movement of the moving body. Then, by obtaining a conversion coefficient from the local coordinate system to the absolute coordinate system and performing coordinate conversion, the acceleration vector in the absolute coordinate system can be detected appropriately. The acceleration vector in the absolute coordinate system detected in this manner is suitable for inertial navigation calculation and the like as described above.

  Further, as a second form, the acceleration detection method according to the first form, wherein the conversion coefficient is obtained by obtaining the conversion coefficient capable of reducing an error that can be included in the detected acceleration vector. An acceleration detection method may be configured.

  According to the second embodiment, a conversion coefficient capable of reducing an error that can be included in the detected acceleration vector is obtained. Therefore, by converting the acceleration vector detected by the acceleration sensor into an acceleration vector in the absolute coordinate system using the conversion coefficient, errors that can be included in the acceleration vector detected by the acceleration sensor can be reduced.

  Further, as a third form, the acceleration detection method according to the first or second form, wherein the adjustment, the conversion coefficient is obtained, and the coordinate conversion is performed, and the moving body is moved. You may comprise the acceleration detection method performed inside.

  According to the third embodiment, the adjustment of the detected posture of the posture sensor, the conversion coefficient, and the coordinate conversion of the detected acceleration vector are performed while the moving body is moving, so that the moving body is moved during the movement. Even when the posture changes, the acceleration vector of the moving body can be detected appropriately.

  Further, as a fourth mode, in the acceleration detection method according to any one of the first to third modes, the adjustment is performed without adjusting the yaw component of the detected posture, and the roll component and You may comprise the acceleration detection method including adjusting at least one of a pitch component.

  According to the fourth aspect, the yaw component in the detected posture is not adjusted, and the adjustment of the detected posture can be easily performed by adjusting at least one of the roll component and the pitch component.

  Further, as a fifth mode, the acceleration detecting method according to any one of the first to fourth modes, wherein the reference vector is measured by a predetermined speed measurement system installed in the mobile unit. The adjusting includes calculating the acceleration vector of the moving body using the velocity vector, and using the detected acceleration vector and the calculated acceleration vector to adjust the detected posture. And adjusting the acceleration.

  According to the fifth aspect, the reference vector includes the velocity vector of the moving object measured by the predetermined velocity measuring system installed on the moving object. In this case, the acceleration vector of the moving object is calculated using the velocity vector of the moving object measured by the speed measurement system. Then, the detected posture of the posture sensor is adjusted using the detected acceleration vector of the acceleration sensor and the calculated acceleration vector. By using the velocity vector of the moving body measured by the velocity measurement system, the detection posture of the posture sensor can be adjusted appropriately.

  More specifically, as in the sixth embodiment, it is possible to configure an acceleration detection method in which the speed measurement system in the fifth embodiment is a vehicle speed detection system or a satellite positioning system.

  According to the sixth embodiment, by using the vehicle speed detection system or the satellite positioning system, it becomes possible to measure the velocity vector of the moving body with high accuracy, and it is possible to use for the calculation of the acceleration vector of the moving body. it can.

  Further, as a seventh aspect, the acceleration detection method of any one of the first to sixth aspects is used to detect an acceleration vector in the absolute coordinate system, and a given movement reference position and the detected acceleration. You may comprise the position calculation method including calculating the position of the said mobile body in the said absolute coordinate system using a vector.

  According to the seventh aspect, the position of the moving object can be calculated with high accuracy by using the given movement reference position and the acceleration vector detected using the acceleration detection method of the above aspect. .

Explanatory drawing of the system configuration | structure of an acceleration detection system. Explanatory drawing of the system configuration | structure of a 1st acceleration detection system. Explanatory drawing of the system configuration | structure of a 2nd acceleration detection system. Explanatory drawing of the system configuration | structure of a 3rd acceleration detection system. Explanatory drawing of the principle of absolute coordinate velocity vector correction | amendment. Explanatory drawing of the system configuration | structure of a navigation system. Explanatory drawing of a function structure of a car navigation apparatus. Explanatory drawing of the system configuration | structure of a 1st position calculation system. The flowchart which shows the flow of a navigation process. Explanatory drawing of the position calculation result by a 1st position calculation system. Explanatory drawing of the attitude | position calculation result by a 1st position calculation system. Explanatory drawing of the system configuration | structure of a 2nd position calculation system. Explanatory drawing of the position calculation result by a 2nd position calculation system. Explanatory drawing of the attitude | position calculation result by a 2nd position calculation system. Explanatory drawing of the system configuration | structure of a 3rd position calculation system.

  Hereinafter, an example of a preferred embodiment of the present invention will be described with reference to the drawings. However, it goes without saying that embodiments to which the present invention can be applied are not limited to the embodiments described below.

1. Principle 1-1. System Configuration FIG. 1 is a system configuration diagram of an acceleration detection system 1 configured by main functional blocks for realizing the acceleration detection method according to the present embodiment. The acceleration detection system 1 is a system that is applied to various types of moving objects, and is a system that detects and outputs accelerations (specifically, acceleration vectors) of the moving objects.

  The acceleration detection system 1 includes an acceleration sensor 3 and an attitude sensor 4 that are a kind of inertial sensors. In FIG. 1, the inertial sensor is illustrated by a double line, and a processing block for performing various processes using the detection result of the inertial sensor is illustrated by a single line to distinguish the two. The same applies to the subsequent drawings.

  The acceleration sensor 3 is a sensor that is installed on a moving body and detects the acceleration of the moving body. For example, a MEMS sensor using a technology of MEMS (Micro Electro Mechanical Systems) is used. In the present embodiment, the acceleration sensor 3 is configured to detect and output a triaxial local coordinate acceleration vector in a triaxial local coordinate system with the moving body as a reference. That is, it is assumed that the acceleration sensor 3 is installed on the moving body in such a posture that the local coordinate system associated with the acceleration sensor 3 matches the coordinate system based on the moving body.

  The attitude sensor 4 is a sensor that detects the attitude of the moving body. For example, the attitude sensor 4 may be configured by only a gyro sensor, or may be configured by further combining a magnetic sensor and an acceleration sensor. The posture sensor 4 is fixed to the moving body and detects the posture of the moving body in a local coordinate system having three orthogonal axes. That is, the posture sensor 4 is a measuring instrument for measuring the posture angle of the moving body, and the value measured by the posture sensor 4 is the direction of the moving body expressed in an absolute coordinate system.

  The acceleration and speed of the moving body have a direction and a magnitude. For this reason, in this specification, a scalar and a vector are appropriately distinguished and described. In principle, simply referring to acceleration and speed means the magnitude of the acceleration and speed (scalar amount), and referring to acceleration vector and speed vector means the acceleration and speed considering the direction and magnitude. Shall. Further, in order to distinguish between a vector defined in the local coordinate system and a vector defined in the absolute coordinate system, a description will be given with the type of the coordinate system added to the head of a word representing each variable.

  In the coordinate conversion, the direction of the acceleration vector among the detected acceleration vectors of the acceleration sensor 3 is coordinate-converted, and the magnitude of the acceleration vector is maintained. In other words, the relative detection direction of the acceleration sensor 3 is coordinate-converted into a direction (absolute direction) in the absolute coordinate system.

  Further, in this specification, the posture angle output from the posture sensor 4 is referred to as “detected posture angle”, and the posture of the moving body determined by the detected posture angle is referred to as “detected posture”. In addition, the posture angle acquired by adjusting the detected posture angle by the posture adjustment unit 20 is referred to as an “adjustment posture angle”, and the posture of the moving body determined by the adjustment posture angle is referred to as an “adjustment posture”.

  The acceleration detection system 1 includes an attitude adjustment unit 20 and a coordinate conversion unit 30 as main processing blocks.

  The posture adjusting unit 20 is an adjusting unit that adjusts the detected posture of the posture sensor 4 using a local coordinate acceleration vector that is a detected acceleration vector of the acceleration sensor 3 and a given reference vector related to the movement of the moving body. .

  The coordinate conversion unit 30 is a conversion unit that converts a local coordinate acceleration vector, which is a detected acceleration vector of the acceleration sensor 3, into an absolute coordinate acceleration vector using the posture adjusted by the posture adjustment unit 20. Specifically, using the adjustment posture that is the detection posture adjusted by the posture adjustment unit 20, a conversion coefficient (specifically, a coordinate conversion matrix) for coordinate conversion from the local coordinate system to the absolute coordinate system of the moving object is obtained. Ask. Then, using the obtained conversion coefficient, the local coordinate acceleration vector is coordinate-converted to an absolute coordinate acceleration vector. As will be described in detail later, the coordinate transformation matrix is a matrix having a transformation coefficient related to coordinate transformation as a component.

  The coordinate conversion unit 30 is a conversion coefficient calculation unit that calculates a conversion coefficient (coordinate conversion matrix) related to coordinate conversion. Since the acceleration vector in the absolute coordinate system is calculated by the coordinate conversion, it can be said that the coordinate conversion unit 30 is an absolute coordinate acceleration vector calculation unit.

1-2. Next, the principle of acceleration detection in this embodiment will be described. First, the detection posture of the posture sensor 4 is adjusted using a local coordinate acceleration vector that is a detection acceleration vector of the acceleration sensor 3 and a given reference vector related to the movement of the moving body. It is one of. Next, a conversion coefficient (coordinate conversion matrix) related to coordinate conversion from the local coordinate system associated with the acceleration sensor 3 to the absolute coordinate system is obtained using the adjusted detection posture. Then, using the obtained conversion coefficient, the acceleration of the moving body is detected by converting the detected acceleration vector of the acceleration sensor 3 into the acceleration vector in the absolute coordinate system.

  As the reference vector related to the movement of the moving body, for example, an acceleration vector or a velocity vector of the moving body can be applied. The reference vector can be obtained by various methods. For example, it can be acquired from a predetermined speed measurement system installed on the moving body. The speed measurement system is a vehicle speed detection system or a satellite positioning system.

  In the vehicle speed detection system, for example, the magnitude of the speed of the moving body can be detected by a vehicle speed pulse. When a vehicle speed detection system is configured by combining a direction sensor such as a magnetic sensor, the moving direction of the moving body can also be detected. In this case, the vehicle speed detection system detects the magnitude and moving direction of the moving body. For this reason, the speed vector of the moving body can be acquired as the reference vector from the vehicle speed detection system.

  The satellite positioning system is typically a GPS (Global Positioning System). In GPS, based on information such as the positions of a plurality of GPS satellites and pseudo distances from the respective GPS satellites to the GPS receiver, position calculation for obtaining the GPS receiver position coordinates and clock error is performed. In addition, since the position calculation using GPS is conventionally well-known, detailed description is abbreviate | omitted here.

  In addition to the position calculation, the GPS can calculate the velocity vector of the GPS receiver by using Doppler generated by the change in the relative positional relationship between the GPS satellite and the GPS receiver. That is, the relative velocity vector between the GPS receiver and the GPS satellite is calculated by measuring the Doppler frequency of the GPS satellite signal received from the GPS satellite. The velocity vector (movement speed and direction) of the GPS satellite can be obtained from satellite orbit data (ephemeris and almanac). Therefore, if the relative velocity vector and the velocity vector of the GPS satellite are used, the velocity vector (movement velocity and movement direction) of the moving body can be calculated. For this reason, the velocity vector of the moving object can be acquired from the GPS as the reference vector.

  If the velocity vector of the moving body is differentiated with respect to time, the acceleration vector of the moving body can be obtained. Therefore, the acceleration vector of the moving body can be calculated by differentiating the velocity vector acquired as a reference vector from the velocity measurement system with respect to time.

  Therefore, in the posture adjustment method of the present embodiment, the acceleration vector of the moving body used for posture adjustment (hereinafter referred to as “posture adjustment acceleration vector”) is calculated using the reference vector. Then, an error included in the detected posture angle (hereinafter referred to as “posture angle error”) is estimated using the local coordinate acceleration vector that is the detected acceleration vector of the acceleration sensor 3 and the calculated posture adjustment acceleration vector. . In this case, it is preferable to calculate an acceleration vector in the absolute coordinate system as the posture adjustment acceleration vector. Finally, the detected posture angle is adjusted using the estimated posture angle error.

  In order to make the description easy to understand, a coordinate system will be specifically exemplified and described. Here, an aircraft coordinate system (Body Frame) (hereinafter referred to as “B frame”) is applied as a local coordinate system, and an NED (North East Down) coordinate system (hereinafter referred to as a northeast lower coordinate system) is known as an absolute coordinate system. , Referred to as “N frame”) will be described as an example.

  The B frame is, for example, an X-axis (roll axis) with the front / rear direction positive with respect to the front of the moving body, a Y-axis (pitch axis) with the right side positive, and a vertical direction with the vertical downward direction positive with Z It is a three-dimensional orthogonal coordinate system with an axis (yaw axis). On the other hand, the N frame is, for example, a three-dimensional orthogonal coordinate system in which the north-south direction with the north as positive is the N axis, the east-west direction with the east as positive is the E axis, and the altitude direction with the vertical downward is the D axis. It is.

The conversion from the B frame acceleration vector to the N frame acceleration vector is formulated by the following equation (1).

“F ^b ” and “f ⁿ ” in equation (1) are a B frame acceleration vector and an N frame acceleration vector, respectively, and are given by the following equations (2) and (3).

Further, “C _b ⁿ ” in the equation (1) is a coordinate conversion matrix from the B frame to the N frame, and is given by the following equation (4).

In Expression (4), (φ, θ, Ψ) respectively represents a rotation angle (roll angle) around the X axis (roll axis), a rotation angle (pitch angle) around the Y axis (pitch axis), and Z The rotation angle (yaw angle) about the axis (yaw axis). Each component of the coordinate transformation matrix “C _b ⁿ ” is a transformation coefficient related to coordinate transformation from the B frame to the N frame.

From the formulas (1) to (4), the following formula (5) is derived.

Here, the expression (5) is regarded as a function of the roll angle “φ” and the pitch angle “θ”, and is formulated as the following expression (6).

  In Expression (6), the posture angles with a hat “^” indicate the actual measurement values of the roll angle “φ” and the pitch angle “θ”, respectively, and the posture angles without the hat “^” are the roll angles “φ”. And the true value of the pitch angle “θ”. The posture angle with a delta “Δ” indicates an error (posture angle error) that can be included in the actual measurement values of the roll angle “φ” and the pitch angle “θ”. “Δφ” is a roll angle error, and “Δθ” is a pitch angle error.

When the equation (6) is Taylor-expanded to the first order term, the following equation (7) is derived.

From the equations (6) and (7), the following equation (8) is derived.

The left side of Equation ^{(8), "f n =} (f x n, f y n, f z n)" in the first term represents the true value of N frame acceleration vector, the second term hat ^ with the “F ⁿ (φ, θ) = (F _x ⁿ (φ, θ), F _y ⁿ (φ, θ), F _z ⁿ (φ, θ))” indicates an actually measured value of the N frame acceleration vector.

N frame the actual measurement value is hat ^ with the acceleration vector ^{"F n (φ, θ)} " is, B-frame acceleration vector "f ^b is a detected acceleration vector of the acceleration sensor ^{_{3 = (f x b, f}} y b, f _z ^b ) ”and the detected attitude angles (φ, θ, Ψ) of the attitude sensor 4 can be calculated by substituting them into the equation (5), respectively. That is, the actual measurement value “F ⁿ (φ, θ)” of the N frame acceleration vector is known.

In contrast, the true value of N frame acceleration vector _{^{"f n = (f x n}} , f y n, f z n)" is unknown, it is impossible to know the value. However, in this embodiment, a reference vector related to the movement of the moving object is given from the outside. Therefore, by using a reference vector, calculates the N frame acceleration vector is an absolute coordinate acceleration vector, which the true value of N frame acceleration vector _{^{"f n = (f x n}} , f y n, f z n)" and Consider and calculate. If the reference vector is sufficiently reliable, sufficient accuracy is guaranteed even if the N frame acceleration vector calculated from the reference vector is regarded as a true value.

  Regarding the right side of Equation (8), the values of each component of the partial differentiation matrix (Jacobi matrix) are obtained by substituting Equation (5) with the measured values of roll angle and pitch angle ((φ, θ) with a hat ^). It is obtained by differentiating. The measured value of the posture angle is the posture angle (detected posture angle) detected by the posture sensor 4.

In Expression (8), for example, the left side is set as “Δρ”, the partial differential matrix on the right side is set as “H”, and the matrix (posture angle error matrix) including the posture angle error on the right side as a component is set as “Δx”. Equation (8) can be rewritten as Equation (9).

The only unknown in equation (9) is the attitude angle error matrix “Δx”. Therefore, the equation (9) can be solved as the following equation (10).

  In the calculation using a computer (computer), the posture angle error matrix “Δx” is approximately obtained. For example, an approximate solution of the attitude angle error matrix “Δx” can be obtained by using a Newton method which is a kind of root finding algorithm and is known as an iterative method.

  When the posture angle error matrix “Δx” is obtained, the posture angle errors (roll angle error “Δφ” and pitch angle error “Δθ”) are detected by the posture angle detected by the posture sensor 4 ((φ, θ, Ψ with a hat ^). )) Is added to adjust the detected posture angle.

  In this embodiment, errors included in the roll angle “φ” and the pitch angle “θ” are approximately obtained, and the roll angle “φ” and the pitch angle “of the detected attitude angles are used by using these errors. Adjust θ ”. That is, the yaw component in the detected posture of the moving body is not adjusted, and only the roll component and the pitch component are adjusted.

  When the coordinate transformation matrix (transformation coefficient) from the local coordinate system to the absolute coordinate system is obtained using the adjustment attitude angle obtained as described above, and the detected acceleration vector of the acceleration sensor 3 is transformed using this, the detected acceleration is obtained. Errors that may be included in the vector are reduced by coordinate transformation. That is, the error is smaller than the error originally included in the detected acceleration vector. The reason is as follows.

  The local coordinate acceleration vector detected by the acceleration sensor 3 may include an error due to movement of the moving body. When the local coordinate acceleration vector that may contain this error is subjected to coordinate conversion, the error is naturally superimposed on the converted absolute coordinate acceleration vector.

  However, as described in the above principle, in this embodiment, the detected posture of the moving body is adjusted using the local coordinate acceleration vector itself detected by the acceleration sensor 3, and the adjustment posture angle that is the adjustment result is used. To obtain a coordinate transformation matrix. This presupposes that an error is included in the detected acceleration vector of the acceleration sensor 3, and in other words, it means obtaining a conversion coefficient that matches the detected acceleration vector. That is, it is an idea to reduce the error included in the detected acceleration vector by including the error in the detected acceleration vector in the error in the detected posture.

  The adjustment of the detected posture (detected posture angle) by the posture adjusting unit 20, the calculation of the conversion coefficient (coordinate conversion matrix) by the coordinate converting unit 30, and the coordinate conversion by the coordinate converting unit 30 are performed while the moving body is moving. It is preferred to do so. Of course, it may be performed when the moving body is stopped, but the method of the present embodiment is particularly effective during the movement of the moving body.

Apart from the method of the present embodiment, a method has been devised that estimates the posture of the moving body using the acceleration vector detected by the acceleration sensor 3 when the moving body is stopped. Specifically, the roll angle “φ” and the pitch angle “θ” are estimated according to the following equations (11) and (12).

  If the moving body is stopped, since the posture of the moving body does not normally change, the posture of the moving body can be correctly estimated by using the equations (11) and (12). Therefore, when the moving body is stopped, a coordinate transformation matrix is obtained using the roll angle “φ” and the pitch angle “θ” estimated according to the equations (11) and (12), and the detected acceleration vector of the acceleration sensor 3 is coordinated. By converting, the absolute coordinate acceleration vector can be obtained appropriately.

  However, the situation is different while the moving body is moving. Due to the movement of the moving body, the posture of the moving body changes at any time. Then, the posture angle estimated at the time of stopping becomes incorrect, and the coordinate transformation matrix obtained from the posture angle estimated at the time of stopping becomes inconsistent with the detected acceleration vector of the acceleration sensor 3. When the detected acceleration vector is coordinate-converted using such a coordinate conversion matrix, the error included in the detected acceleration vector cannot be reduced, and a large error remains in the absolute coordinate acceleration vector.

  In that respect, in the method of the present embodiment, since a coordinate transformation matrix that matches the detected acceleration vector of the acceleration sensor 3 is obtained, there is no problem even if the posture of the moving body changes at any time during movement. Thus, an absolute coordinate acceleration vector in which the influence of the error is suppressed can be obtained.

  The acceleration detection system 1 shown in FIG. 1 shows the minimum components for realizing acceleration detection in the present embodiment. Actually, various variations of the acceleration detection system are conceivable. Therefore, variations of the acceleration detection system will be described below. In addition, about the functional block same as the acceleration detection system 1, the same code | symbol is attached | subjected and description is abbreviate | omitted and it demonstrates centering on a different part from the acceleration detection system 1. FIG.

  FIG. 2 is an explanatory diagram of the first acceleration detection system 1A. The first acceleration detection system 1 </ b> A includes a posture adjustment acceleration vector calculation unit 40 in addition to the functional blocks of the acceleration detection system 1. The posture adjustment acceleration vector calculation unit 40 calculates a posture angle adjustment acceleration vector (posture angle adjustment acceleration vector) using a reference vector related to the movement of the moving object.

  Specifically, the posture adjustment acceleration vector calculation unit 40 calculates the absolute coordinate acceleration vector using the reference velocity vector of the moving body externally input as the reference vector. For example, when an absolute coordinate velocity vector is input as the reference velocity vector, the absolute coordinate acceleration vector is calculated by differentiating the absolute coordinate velocity vector with respect to time. Then, the calculated absolute coordinate acceleration vector is output to the posture adjustment unit 20 as a posture adjustment acceleration vector.

  FIG. 3 is an explanatory diagram of the second acceleration detection system 1B. The second acceleration detection system 1B includes an attitude adjustment unit 20, a coordinate conversion unit 30, an attitude adjustment acceleration vector calculation unit 40, a velocity vector calculation unit 50, and a velocity vector correction unit 60 as processing blocks. .

  The velocity vector calculation unit 50 calculates the velocity vector (absolute coordinate velocity vector) of the moving body in the absolute coordinate system by integrating and adding the absolute coordinate acceleration vector input from the coordinate conversion unit 30. Then, the calculated absolute coordinate velocity vector is output to the velocity vector correction unit 60. Here, “integration” means that values for a predetermined time are cumulatively added. “Adding” means adding the result obtained by integration to the result of the last update.

  Specifically, the velocity vector calculation unit 50 integrates the absolute coordinate acceleration vector for a predetermined time to calculate the change in the velocity vector of the moving body during the predetermined time. Then, the absolute coordinate velocity vector is newly calculated and updated by adding the calculated change in the velocity vector to the last updated absolute coordinate velocity vector.

  The velocity vector correction unit 60 corrects the absolute coordinate velocity vector input from the velocity vector calculation unit 50. Specifically, the absolute coordinate velocity vector is corrected using a reference velocity vector input externally as a reference vector related to the movement of the moving object. As the reference velocity vector, for example, a velocity vector in an absolute coordinate system can be acquired. For this reason, for example, the absolute coordinate velocity vector is corrected by averaging two types of absolute coordinate velocity vectors.

  As an averaging process in this case, a simple arithmetic average can be applied, or a weighted average can be applied. If the reference speed vector is trusted and the speed vector is corrected, the weight of the reference speed vector may be set higher than the weight of the speed vector calculated by the speed vector calculation unit 50 and weighted. Conversely, if the speed vector calculated by the speed vector calculation unit 50 is trusted and the speed vector is corrected, the weight of the speed vector may be set higher than the weight of the reference speed vector and weighted averaged.

  The velocity vector correction unit 60 externally outputs an absolute coordinate velocity vector (hereinafter referred to as “absolute coordinate correction velocity vector”) obtained as a correction result, and feeds it back to the posture adjustment acceleration vector calculation unit 40. The posture adjustment acceleration vector calculation unit 40 calculates a posture adjustment acceleration vector using the absolute coordinate correction velocity vector input from the velocity vector correction unit 60, and outputs it to the posture adjustment unit 20.

  FIG. 4 is an explanatory diagram of the third acceleration detection system 1C. The third acceleration detection system 1C is a modification of the second acceleration detection system 1B. Specifically, unlike the second acceleration detection system 1B, in the third acceleration detection system 1C, the detected attitude of the attitude sensor 4 is input to the attitude adjustment unit 20 and the velocity vector correction unit 60, respectively. It is configured.

  The velocity vector correction unit 60 corrects the absolute coordinate velocity vector input from the velocity vector calculation unit 50 using the reference velocity externally input as the reference information related to the movement of the moving body and the detected posture of the posture sensor 4. For example, when reference information can be acquired from the vehicle speed detection system, if the vehicle speed detection system is configured to detect only the speed of the moving body, the speed information is acquired as the reference information, and the attitude The absolute coordinate velocity vector is corrected in combination with the detection posture of the sensor 4.

  FIG. 5 is a diagram for explaining the principle of correcting the absolute coordinate velocity vector. Here, N frames are considered as an absolute coordinate system, and a case where a velocity vector is corrected in a two-dimensional plane composed of an east-west direction (E axis) and a north-south direction (N axis) will be illustrated and described. In FIG. 5, the horizontal axis indicates the east-west axis (E axis), and the vertical axis indicates the north-south axis (N axis).

  When the east-west direction component and the north-south direction component of the N frame acceleration vector are integrated and added to the last updated N frame velocity vector, the east-west direction component and the north-south direction component of the N frame velocity vector are obtained. In FIG. 5, the arrow described on the E axis indicates the east-west direction component of the N frame rate vector, and the arrow described on the N axis indicates the north-south direction component of the N frame rate vector.

  By combining the east-west direction component and the north-south direction component, an N frame velocity vector having a magnitude of “| V1 |” and a direction of “θ1” is obtained. That is, the N frame velocity vector is defined by a circular coordinate having a moving radius of “| V1 |” and a declination of “θ1”. However, in FIG. 5, the direction of the N frame velocity vector is represented by an angle formed by the velocity vector and the E axis. The N frame rate vector is the N frame rate vector calculated by the rate vector calculating unit 50.

  The magnitude and direction of the N frame rate vector are corrected as follows, for example. First, the N frame rate vector magnitude “| V1 |” (first rate) and the externally input reference rate (second rate) are averaged and combined. The speed obtained as a result of the synthesis is assumed to be the magnitude “| V2 |” of the N frame correction speed vector.

  Next, the direction “θ1” of the N frame velocity vector is corrected based on the detected posture of the posture sensor 4 so as to match the moving direction “θ2” of the moving body. The direction “θ1” of the N frame velocity vector is a direction obtained indirectly by integrating and adding the N frame corrected acceleration vector, whereas the moving direction “θ2” is directly determined from the detected posture of the posture sensor 4. This is the direction of the moving object. Therefore, by the above correction, the direction of the N frame correction speed vector becomes a more accurate direction along the moving direction of the moving body.

2. Example Next, an example of a position calculation system and a navigation system to which the above-described acceleration detection system is applied will be described. However, it goes without saying that the embodiments to which the present invention can be applied are not limited to the embodiments described below.

2-1. System Configuration FIG. 6 is an explanatory diagram of the system configuration of the navigation system in the present embodiment. In the navigation system, a car navigation device 1000, which is a type of electronic device including an acceleration detection device and a position calculation device, is mounted on a four-wheeled vehicle (hereinafter simply referred to as "automobile") that is a type of mobile body. System.

  The car navigation apparatus 1000 is an electronic device mounted on a car, and includes an acceleration detection device that calculates an acceleration vector of the car and a position calculation device that calculates the position of the car. In addition, the car navigation apparatus 1000 includes an IMU (Inertial Measurement Unit) 600 as a sensor unit having the acceleration sensor 3 and the gyro sensor 5.

  The IMU 600 is a sensor unit known as an inertial measurement unit, and is designed to detect and output an acceleration vector and an angular velocity of a coordinate system relative to the automobile. In the present embodiment, the car navigation apparatus 1000 will be described as being installed and fixed in an automobile in such a posture that the relative coordinate system coincides with the B frame.

  In addition, the car navigation apparatus 1000 includes a GPS unit 200 as a sensor unit using GPS, which is a kind of satellite positioning system. The GPS unit 200 is configured to be able to measure and output, for example, the position and speed vector of an automobile in an N frame using a GPS satellite signal transmitted from a GPS satellite.

  The car navigation apparatus 1000 calculates the position of the automobile by performing inertial navigation calculation processing using the detection result of the IMU 600. Then, a navigation screen is generated by performing a map matching process on the calculated position, and the generated navigation screen is displayed on the display as the display unit 500.

2-2. Functional Configuration FIG. 7 is a block diagram illustrating an example of a functional configuration of the car navigation apparatus 1000. The car navigation apparatus 1000 includes a GPS antenna 100, a GPS unit 200, a processing unit 300, an operation unit 400, a display unit 500, an IMU 600, and a storage unit 700 as main functional configurations. .

  The GPS antenna 100 is an antenna that receives an RF (Radio Frequency) signal including a GPS satellite signal transmitted from a GPS satellite that is a type of positioning satellite, and outputs a received signal to the GPS unit 200. The GPS satellite signal is a 1.57542 [GHz] communication signal modulated by a CDMA (Code Division Multiple Access) system known as a spread spectrum system by a CA (Coarse and Acquisition) code which is a kind of spreading code. The CA code is a pseudo-random noise code having a repetition period of 1 ms with a code length of 1023 chips as 1 PN frame, and is different for each GPS satellite.

  The GPS unit 200 is a circuit or device that measures the position, moving speed, and moving direction of an automobile based on a signal output from the GPS antenna 100, and is a functional block corresponding to a so-called GPS receiving device. Although not shown, the GPS unit 200 includes an RF receiving circuit unit and a baseband processing circuit unit. Note that the RF receiving circuit unit and the baseband processing circuit unit can be manufactured as separate LSIs (Large Scale Integration) or can be manufactured as one chip.

  The baseband processing circuit unit performs correlation processing or the like on the received signal output from the RF receiving circuit unit to capture and track the GPS satellite signal. Then, based on satellite orbit data, time data, and the like extracted from the GPS satellite signal, a predetermined position calculation is performed to calculate the position of the automobile. Further, the baseband processing circuit unit performs a predetermined speed calculation calculation based on the Doppler frequency of the GPS satellite signal received from the GPS satellite to calculate the speed vector of the automobile. In this embodiment, the GPS unit 200 will be described as being configured to be able to calculate and output the position (position coordinates) and speed vector of an automobile in N frames.

  The processing unit 300 is a control device that comprehensively controls each unit of the car navigation apparatus 1000 according to various programs such as a system program stored in the storage unit 700, and includes a processor such as a CPU (Central Processing Unit). Composed.

  The processing unit 300 performs inertial navigation calculation processing using the detection result of the IMU 600 to calculate the position (position coordinates) of the automobile. And based on the calculated position, the process which displays the map which pointed to the present position of a motor vehicle on the display part 500 is performed.

  FIG. 8 is a diagram illustrating an example of a system configuration of a first position calculation system 2A that is a position calculation system to which the first acceleration detection system 1A described in the principle is applied, and the processing unit 300 is represented as a functional block. It is a conceptual diagram.

  The processing unit 300 includes a posture calculation unit 310, a posture adjustment unit 320, a coordinate conversion unit 330, a posture adjustment acceleration vector calculation unit 340, a velocity vector calculation unit 350, and a position calculation unit 380 as functional units. . The posture adjustment unit 320 to the velocity vector calculation unit 350 correspond to the posture adjustment unit 20 to the velocity vector calculation unit 50 of the acceleration detection system described in the principle.

  The attitude calculation unit 310 calculates the attitude angle of the automobile by integrating and adding the B frame angular velocity detected by the gyro sensor 5. In other words, the gyro sensor 5 and the attitude calculation unit 310 constitute the attitude sensor 4 in the acceleration detection system described in the principle.

  Specifically, the posture calculation unit 310 calculates the change direction and the change amount of the posture angle of the moving body during the predetermined time by integrating the B frame angular velocity for a predetermined time detected by the gyro sensor 5. To do. The predetermined time is set longer than the detection time interval of the B frame angular velocity, and is usually set to a fixed time.

  For example, the predetermined time is set to 100 milliseconds, and the detection time interval of the B frame angular velocity (more precisely, the sampling time interval of the signal output from the gyro sensor 5) is appropriately adjusted between 1 to 10 milliseconds. In this case, the change direction and change amount of the posture angle of the moving body are calculated every 100 milliseconds regardless of the detection time interval. Then, the latest posture angle is calculated by adding the change direction and change amount of the posture angle calculated every 100 milliseconds to the last updated posture angle.

  In the following description, the attitude angle calculated by the attitude calculation unit 310 is referred to as “calculated attitude angle”, and the attitude of the vehicle determined by the calculated attitude angle is referred to as “calculated attitude”. The posture calculation unit 310 outputs the calculated posture (calculated posture angle) to the posture adjustment unit 320.

  The posture adjustment acceleration vector calculation unit 340 acquires an N frame GPS velocity vector from the GPS unit 200 as a reference velocity vector, and calculates a posture adjustment acceleration vector. Then, the calculated attitude adjustment acceleration vector is output to attitude adjustment section 320.

  The position calculation unit 380 calculates the position of the vehicle in the N frame using the vehicle N frame speed vector calculated by the speed vector calculation unit 350. Specifically, the position change of the vehicle is calculated by integrating and adding the N frame speed vector. Then, the position of the automobile is calculated / updated by cumulatively adding the calculated position changes with the given movement reference position (initial position) of the automobile as a reference. The movement reference position (initial position) can be, for example, a position input by the user as a navigation start position.

  Returning to the description of FIG. 7, the operation unit 400 is an input device configured by, for example, a touch panel or a button switch, and outputs a pressed key or button signal to the processing unit 300. By operating the operation unit 400, various instructions such as destination input are input.

  The display unit 500 is a display device that includes an LCD (Liquid Crystal Display) or the like and performs various displays based on a display signal input from the processing unit 300. A navigation screen or the like is displayed on the display unit 500.

  The IMU 600 includes an acceleration sensor 3 and a gyro sensor 5, and is configured to be able to detect an acceleration vector in each axis of the B frame and an angular velocity around each axis. The acceleration sensor 3 and the gyro sensor 5 may be independent sensors or may be integrated sensors.

  The storage unit 700 is configured by a storage device such as a ROM (Read Only Memory), a flash ROM, or a RAM (Random Access Memory), and various types for realizing various functions such as a system program of the car navigation device 1000 and a navigation function. Stores programs, data, etc. In addition, it has a work area for temporarily storing data being processed and results of various processes.

2-3. Data Configuration As shown in FIG. 7, the storage unit 700 stores a navigation program 701 that is read by the processing unit 300 and executed as a navigation process (see FIG. 9) as a program. The navigation process will be described later in detail using a flowchart.

  The storage unit 700 also includes sensor data 703, reference data 705, calculated attitude angle data 707, attitude angle error data 709, adjusted attitude angle data 711, movement vector data 713, and calculated positions. Data 715 is stored.

  The sensor data 703 is data in which the B frame acceleration vector and the B frame angular velocity detected by the IMU 600 are stored.

  The reference data 705 is data in which a reference value related to the movement of the automobile is stored. In this embodiment, the N frame speed vector of the automobile measured by the GPS unit 200 is stored as the reference data 705.

  The calculated attitude angle data 707 is data in which the attitude angle of the automobile calculated by the attitude calculation unit 310 is stored. The attitude angles of the roll component, pitch component, and yaw component are calculated and stored.

  The posture angle error data 709 is data in which the posture angle error estimated by the posture adjustment unit 320 performing the posture angle error estimation process is stored. For example, the roll component and pitch component errors are calculated and stored.

  The adjusted attitude angle data 711 is data in which the adjusted attitude angle of the automobile acquired by the attitude adjustment unit 320 adjusting the calculated attitude angle is stored.

  The movement vector data 713 is vector data indicating the moving state of the automobile, and stores data such as an N frame acceleration vector calculated by the coordinate conversion unit 330 and an N frame velocity vector calculated by the velocity vector calculation unit 350. The

  The calculated position data 715 is data in which the position of the automobile calculated by the position calculation unit 380 is stored.

2-4. Processing Flow FIG. 9 is a flowchart showing the flow of navigation processing executed in the car navigation device 1000 when the navigation program 701 stored in the storage unit 700 is read and executed by the processing unit 300. . In the following navigation processing, it is assumed that the acceleration vector and angular velocity detected by the IMU 600 are stored in the sensor data 703 of the storage unit 700 as needed.

  First, the posture calculation unit 310 performs a posture angle calculation process (step A1). Specifically, the attitude angle of the vehicle is calculated by integrating and adding the detection results of the gyro sensor 5 and stored in the storage unit 700 as calculated attitude angle data 707.

  Next, the processing unit 300 determines whether or not the reference data 705 has been acquired from the GPS unit 200 (step A3). If it is determined that the reference data 705 has not been acquired (step A3; No), the processing unit 300 proceeds to step A9. And migrate the process.

  If it is determined that the reference data 705 has been acquired (step A3; Yes), the posture adjustment acceleration vector calculation unit 340 performs posture adjustment acceleration vector calculation processing (step A5). That is, the N-frame GPS velocity vector acquired from the GPS unit 200 is time-differentiated to calculate the attitude adjustment acceleration vector in the N frame.

  Next, the posture adjustment unit 320 performs posture angle error estimation processing (step A7). Specifically, using the B frame acceleration vector detected by the acceleration sensor 3, the posture adjustment acceleration vector calculated in step A5, and the posture angle calculated in step A1, for example, iterative calculation using the Newton method. To estimate the attitude angle error. Then, the estimated posture angle error is stored in the storage unit 700 as posture angle error data 709.

  Thereafter, the posture adjustment unit 320 adjusts the latest calculated posture angle stored in the calculated posture angle data 707 using the latest posture angle error stored in the posture angle error data 709, and serves as adjusted posture angle data 711. It memorize | stores in the memory | storage part 700 (step A9).

  Next, the coordinate conversion unit 330 performs an N frame acceleration vector calculation process (step A11). Specifically, a coordinate conversion matrix from the B frame to the N frame is obtained using the latest adjustment attitude angle stored in the adjustment attitude angle data 711. Then, the B frame acceleration vector detected by the acceleration sensor 3 is converted into an N frame acceleration vector using the coordinate conversion matrix, and is stored in the storage unit 700 as movement vector data 713.

  Thereafter, the speed vector calculation unit 350 performs an N frame speed vector calculation process (step A13). That is, the N frame acceleration vector for a predetermined time is integrated and added to the latest N frame velocity vector to newly calculate the N frame velocity vector. Then, it is stored in the storage unit 700 as movement vector data 713.

  Next, the position calculation unit 380 performs position calculation processing (step A15). Specifically, the position change of the moving body is calculated by integrating and adding N frame velocity vectors for a predetermined time stored in the movement vector data 713. The calculated position change is added to the latest calculated position stored in the calculated position data 715 to newly calculate / update the position of the automobile (hereinafter referred to as “inertial navigation calculation position”). To do.

  In addition, in the position calculation process of step A15, it is also possible to acquire the GPS calculated position from the GPS unit 200, and to determine the final position of the automobile (final position) in consideration of the acquired GPS calculated position. It is. For example, an average process between the inertial navigation calculation position and the GPS calculation position may be performed. Further, the inertial navigation calculation position may be reset at the GPS calculation position at regular timing. In this case, the inertial navigation calculation is newly started with the GPS calculated position as the movement reference position.

  Thereafter, the position calculation unit 380 performs a map matching process on the final position calculated in step A15, and updates the navigation screen of the display unit 500 (step A17). Since the process related to map matching is conventionally known, detailed description thereof is omitted here.

  Next, the processing unit 300 determines whether or not to end the navigation process (step A19). For example, when a navigation end instruction operation is performed by the user via the operation unit 400, it is determined that the navigation process is to be ended. When it determines with not complete | finishing a process yet (step A19; No), the process part 300 returns to step A1. If it is determined that the process is to be ended (step A19; Yes), the processing unit 300 ends the navigation process.

2-5. Experimental Results FIG. 10 is a diagram illustrating an example of experimental results obtained by actually performing position calculation in the above-described first position calculation system 2A. The automobile was caused to make one turn along a predetermined circulation route, and the positions calculated in that case were plotted on a two-dimensional plane of east, west, south, and north.

  In FIG. 10, the horizontal axis indicates the east-west direction, and the vertical axis indicates the north-south direction (unit: meters). The starting point was 0m in the east-west direction and 0m in the north-south direction, and the route followed the route from the departure point to the north direction and one round in the counterclockwise direction. The goal point is the same as the starting point. In addition, the true trajectory of the car is indicated by a dotted line, the trajectory obtained by determining the car position without adjusting the attitude angle is indicated by a thin solid line, and the trajectory obtained by adjusting the attitude angle and obtaining the car position is indicated by a thick solid line. Show.

  As can be seen from this figure, when the position of the car is obtained without adjusting the attitude angle, a position along the true trajectory is obtained at first, but it gradually increases with time. It can be seen that a large positional deviation has finally occurred. This is because an error is included in the attitude angle of the vehicle calculated by the attitude calculation unit 310, and when the B frame acceleration vector is coordinate-converted using the attitude angle, the error is superimposed on the N frame acceleration vector. It is because it ends.

  On the other hand, when the position of the automobile is obtained by adjusting the attitude angle, it can be seen that a result along a substantially true locus is obtained. Even if time passes, it does not deviate greatly from the true trajectory, and it can be seen that accumulation of errors with the passage of time is prevented.

  FIG. 11 is a graph showing the time change of the posture angle in the position calculation. 11A shows the time change of the roll angle “φ”, FIG. 11B shows the time change of the pitch angle “θ”, and FIG. 11C shows the time change of the yaw angle “Ψ”. .

  In each figure, the true attitude angle of the automobile is indicated by a dotted line. The true values of the roll angle “φ” and the pitch angle “θ” are zero. In addition, a posture angle that has not been adjusted (calculated posture angle) is indicated by a thin solid line, and a posture angle that has been adjusted (adjusted posture angle) is indicated by a thick solid line. In the present embodiment, since the yaw component is not adjusted, the adjustment results are shown only for the roll component and the pitch component.

  As can be seen from the results, for the posture angle that is not adjusted, both the roll angle “φ” and the pitch angle “θ” tend to deviate from the true value with time. . It is characteristic that the amplitude of the deviation amount (posture angle error) from the true value of the posture angle increases with time. As shown in FIG. 10, the error in the calculated attitude angle is coupled with the error in the detected acceleration vector and is a factor that greatly reduces the accuracy of position calculation.

  On the other hand, with regard to the adjustment posture angle, it can be seen that both the roll angle “φ” and the pitch angle “θ” have values that are substantially in line with the true value even if time passes. As is clear from the results of FIG. 10, the accuracy of position calculation is obtained by performing coordinate conversion of detected acceleration using the attitude angle adjusted by the method of the present embodiment, and calculating the position by performing inertial navigation calculation. Will improve.

2-6. Effects In the acceleration detection system 1A, the posture calculation unit 310 calculates the posture (posture angle) of the vehicle using the B frame angular velocity detected by the gyro sensor 5 installed in the vehicle. The attitude adjustment unit 320 adjusts the attitude of the vehicle calculated by the attitude calculation unit 310 using the B frame acceleration vector detected by the acceleration sensor 3 installed in the vehicle and the reference velocity vector measured by the GPS unit 200. . That is, the attitude adjustment acceleration vector calculation unit 340 calculates the N frame acceleration vector of the automobile using the reference speed vector. The posture adjustment unit 320 estimates a posture angle error by performing a predetermined estimation calculation for estimating a posture angle error using the B frame acceleration vector and the N frame acceleration vector as input values. Then, posture adjustment unit 320 adjusts the posture angle using the estimated posture angle error.

  The acceleration vector detected by the acceleration sensor 3 is an acceleration vector in a local coordinate system (for example, B frame) associated with the acceleration sensor 3. On the other hand, for example, in inertial navigation calculation, the calculation is generally performed based on an absolute coordinate system (for example, N frame) that defines the moving space of the moving body. Since the detection result of the gyro sensor 5 may include an error, an error may also be included in the attitude angle of the automobile calculated using this. Therefore, the attitude angle of the vehicle calculated using the detection result of the gyro sensor 5 is adjusted using the detected acceleration vector of the acceleration sensor 3 and the acceleration vector obtained from the measurement result of the GPS unit 200. Then, by obtaining a coordinate transformation matrix from the local coordinate system (B frame) to the absolute coordinate system (N frame), the acceleration vector in the absolute coordinate system can be obtained appropriately.

  The coordinate conversion unit 330 uses the posture angle adjusted by the posture adjustment unit 320 to obtain a coordinate conversion matrix from the B frame to the N frame. This coordinate transformation matrix is a coordinate transformation matrix that can reduce an error that can be included in the B frame acceleration vector detected by the acceleration sensor 3 by coordinate transformation. That is, by converting the B frame acceleration vector using the coordinate conversion matrix defined based on the adjustment attitude angle, the error that can be included in the B frame acceleration vector is reduced, and the influence of the error is suppressed. You can get a vector.

  The velocity vector calculation unit 350 calculates an N frame velocity vector using the N frame acceleration vector calculated by the coordinate conversion unit 330. Since the N frame velocity vector is obtained using the N frame acceleration vector in which the error is reduced, it is possible to prevent the error included in the B frame acceleration vector detected by the acceleration sensor 3 from being superimposed on the N frame velocity vector. Accumulation of errors with progress can be prevented.

  In the acceleration detection method of the present embodiment, the detected acceleration vector is converted into coordinates by converting a detected acceleration vector by obtaining a conversion coefficient (conversion matrix) that can reduce errors that can be included in the detected acceleration vector of the acceleration sensor 3 by coordinate conversion. The acceleration vector of the absolute coordinate system in which the influence of the error is suppressed can be detected without requiring correction of itself. And the position of a moving body can be calculated | required more correctly by using the given movement reference position and the acceleration vector in the detected absolute coordinate system.

3. Modifications Embodiments to which the present invention can be applied are not limited to the above-described embodiments, and can be changed as appropriate without departing from the spirit of the present invention. Hereinafter, although a modification is demonstrated, the same code | symbol is attached | subjected about the component same as said Example, description is abbreviate | omitted, and it demonstrates centering on a different part from said Example.

3-1. Position calculation system Of course, it is possible to construct a position calculation system by applying the acceleration detection system of various variations described in the principle.

  FIG. 12 is a diagram illustrating an example of a system configuration of a second position calculation system 2B to which the second acceleration detection system 1B is applied. In the second position calculation system 2B, the processing unit 300 includes an attitude calculation unit 310, an attitude adjustment unit 320, a coordinate conversion unit 330, an attitude adjustment acceleration vector calculation unit 340, a speed vector calculation unit 350, a speed A vector correction unit 360 and a position calculation unit 380 are provided as functional blocks.

  The speed vector correction unit 360 acquires an N frame GPS speed vector as a reference speed vector from the GPS unit 200 and corrects the N frame speed vector calculated by the speed vector calculation unit 350. Then, the N frame correction speed vector obtained as the correction result is output to position calculation section 380. Further, the N frame correction speed vector is fed back to the attitude adjustment acceleration vector calculation unit 340.

  FIG. 13 is a diagram illustrating an example of an experimental result of actual position calculation using the second position calculation system 2B. The way of viewing the figure is the same as in FIG. As can be seen from the figure, even when the position calculation is performed using the second position calculation system 2B, a trajectory along the true trajectory of the automobile is obtained, and accumulation of errors is suppressed. Recognize.

  FIG. 14 is a graph showing the temporal change of the posture angle corresponding to the position calculation result of FIG. The way of viewing the figure is the same as in FIG. It can be seen that even when the posture angle is adjusted by the second position calculation system 2B, an increase in the posture angle error with the passage of time is suppressed, and a posture angle close to the true posture angle is obtained.

  FIG. 15 is a diagram illustrating an example of a system configuration of a third position calculation system 2C to which the third acceleration detection system 1C is applied. In the third position calculation system 2C, the vehicle attitude (calculated attitude) calculated by the attitude calculation unit 310 is input to the speed vector correction unit 360, and the vehicle speed detected by the vehicle speed detection system 800 is the speed vector. It is configured to be input to the correction unit 360.

  The vehicle speed detection system 800 is a system that detects the speed of an automobile, and is configured to include vehicle speed pulses, for example. The vehicle speed detection system 800 detects the vehicle speed by generating a voltage or pulse corresponding to the rotational speed of the axle.

  The speed vector correction unit 360 corrects the N frame speed vector calculated by the speed vector calculation unit 350 using the calculated posture input from the posture calculation unit 310 and the vehicle speed input from the vehicle speed detection system 800. Then, an N frame correction speed vector, which is the correction result, is output to position calculation section 380.

3-2. Attitude Sensor In the above embodiment, the gyro sensor is exemplified as the attitude sensor, but the attitude sensor is not limited to this. For example, the orientation of the moving body may be detected using an orientation sensor known as a geomagnetic sensor. Further, a posture sensor may be configured by combining a gyro sensor with an orientation sensor or an acceleration sensor.

3-3. Coordinate System In the above embodiment, the machine coordinate system (B frame) is given as an example of the local coordinate system, and the northeast lower coordinate system (N frame) is given as an example of the absolute coordinate system. However, these coordinate systems are merely examples, and it is needless to say that other coordinate systems may be applied. For example, an ENU (East North Up) coordinate system known as the Tohoku upper coordinate system or an ECEF (Earth Centered Earth Fixed) coordinate system known as the earth-centered earth fixed coordinate system may be applied as the absolute coordinate system.

3-4. Adjustment of detected posture In the above-described embodiment, it has been described that both the roll component and the pitch component of the detected posture of the posture sensor are adjusted. However, only one of the roll component and the pitch component is adjusted. It is good. For example, when adjusting only the roll component, Equation (8) may be rewritten to the equation of the roll angle “φ” and the roll angle error “Δφ” may be estimated. Further, when adjusting only the pitch component, the formula (8) may be rewritten with the formula of the pitch angle “θ” and the pitch angle error “Δθ” may be estimated.

3-5. Attitude Angle Error Estimation Calculation The attitude angle error estimation calculation described in the above embodiment is merely an example, and the design can be changed as appropriate. For example, in the above embodiment, the posture angle error is estimated based on the equations (8) to (10) using the Newton method which is a kind of root finding algorithm. Of course, as the root finding algorithm, in addition to the Newton method, algorithms such as the secant method and the pinch-shot method can be applied.

  In addition, unlike the above-described embodiment, the posture angle error can be estimated using a Kalman filter. In this case, for example, the posture angle error is set as a state estimated value (state vector), and the difference between the detected acceleration vector of the acceleration sensor and the posture adjustment acceleration vector obtained from the reference vector related to the movement of the moving object is an observed value. What is necessary is just to set as (observation vector). Then, a prediction calculation (time update) and a correction calculation (observation update) are repeatedly performed to estimate an optimum state estimated value (attitude angle error).

3-6. Electronic Device In the above embodiment, the case where the present invention is applied to a navigation device mounted on a four-wheeled vehicle has been described as an example. However, an electronic device to which the present invention can be applied is not limited thereto. For example, the present invention may be applied to a navigation device mounted on a two-wheeled vehicle, or may be applied to a portable navigation device.

  Of course, the present invention can be similarly applied to electronic devices for purposes other than navigation. For example, the present invention can be similarly applied to other electronic devices such as a mobile phone, a personal computer, and a PDA (Personal Digital Assistant), and acceleration detection and position calculation of the electronic device can be realized.

3-7. Satellite positioning system In the above-described embodiment, the GPS has been described as an example of the satellite positioning system. Of course, other satellite positioning systems may be used.

DESCRIPTION OF SYMBOLS 1 Acceleration detection system, 2 Position calculation system, 3 Acceleration sensor, 4 Attitude sensor, 5 Gyro sensor, 20 Attitude adjustment part, 30 Coordinate conversion part, 40 Attitude adjustment acceleration vector calculation part, 50 Speed vector calculation part, 60 Speed vector Correction unit, 100 GPS antenna, 200 GPS unit, 300 processing unit, 400 operation unit, 500 display unit, 600 IMU, 700 storage unit

Claims (8)

Adjusting the detected attitude of the attitude sensor installed in the moving body using a detected acceleration vector of the acceleration sensor installed in the moving body and a given reference vector related to the movement of the moving body;
Using the adjusted detection attitude to obtain a conversion coefficient for coordinate conversion from a local coordinate system associated with the acceleration sensor to an absolute coordinate system;
Using the conversion coefficient, coordinate-transforming the detected acceleration vector into an acceleration vector in the absolute coordinate system;
An acceleration detection method including: Obtaining the conversion coefficient is to obtain the conversion coefficient capable of reducing an error that may be included in the detected acceleration vector.
The acceleration detection method according to claim 1. Performing the adjustment, obtaining the conversion coefficient, and converting the coordinates during the movement of the moving body;
The acceleration detection method according to claim 1 or 2. The adjusting includes adjusting at least one of a roll component and a pitch component without adjusting the yaw component of the detected posture,
The acceleration detection method as described in any one of Claims 1-3. The reference vector includes a velocity vector of the moving body measured by a predetermined speed measuring system installed on the moving body,
The adjustment is
Calculating an acceleration vector of the moving object using the velocity vector;
Adjusting the detected posture using the detected acceleration vector and the calculated acceleration vector;
including,
The acceleration detection method as described in any one of Claims 1-4. The speed measurement system is a vehicle speed detection system or a satellite positioning system.
The acceleration detection method according to claim 5. Detecting an acceleration vector in the absolute coordinate system using the acceleration detection method according to claim 1;
Calculating a position of the moving body in the absolute coordinate system using a given movement reference position and the detected acceleration vector;
Position calculation method including An adjustment unit that adjusts the detected posture of the posture sensor installed in the moving body using a detected acceleration vector of the acceleration sensor installed in the moving body and a given reference vector related to the movement of the moving body;
A conversion coefficient calculation unit for obtaining a conversion coefficient related to coordinate conversion from the local coordinate system to the absolute coordinate system associated with the acceleration sensor, using the adjusted detection posture;
A conversion unit that converts the detected acceleration vector into an acceleration vector in the absolute coordinate system using the conversion coefficient;
Acceleration detection apparatus comprising:

Images