CN103471615A - Method for quickly detecting failure of dual-redundancy inertial navigation system - Google Patents

Abstract

The invention belongs to the field of inertia navigation techniques, and particularly relates to a method for quickly detecting a failure of a dual-redundancy inertial navigation system. The method disclosed by the invention comprises the following steps of detecting a bottom layer inertia navigation failure; detecting a failure in navigation software; and detecting a failure in interface control software. According to the method provided by the invention, the technical problems of low detection efficiency and difficulty in location of a specific fault device in the method for quickly detecting the failure of the dual redundancy inertial navigation system in the prior art are solved, inertia elements and key signals of two sets of inertial navigation systems in a local datum can be detected in real time, the detection efficiency is high, the fault elements can be accurately located, and the guarantee is provided for normal work of the local datum.

Description

Rapid fault detection method for dual-redundancy inertial navigation system

Technical Field

The invention belongs to the technical field of inertial navigation, and particularly relates to a quick fault detection method for a dual-redundancy inertial navigation system.

Background

In the modern aerospace industry, the reliability of a navigation and guidance system is a 'short plate' of the weapon system, and how to improve the reliability of an inertial navigation system is a difficult problem in weapon system engineering application.

The following two methods are generally used to improve the reliability of inertial navigation systems: firstly, the reliability of an accelerometer and a gyroscope which are key components of the system is improved; and secondly, a high-reliability system is constructed by adopting a redundancy design and adding redundant sensors. In the two methods, the reliability of a single accelerometer and a single gyroscope is the most fundamental method, but because of the limitation of factors such as technical level, expenditure and the like, the reliability of the inertial navigation system is difficult to be enhanced by simply improving the performance of an inertial element; the reliability of the inertial navigation system is improved by a redundancy technology, which is an economic and feasible technical means.

At present, most of redundancy technologies adopted at home and abroad are system-level redundancy, namely, a scheme that two or more sets of inertial navigation systems are mutually hot-backed is adopted, and when a certain set of inertial navigation system fails, navigation information of the other set of inertial navigation system is directly switched and output. The system level redundancy has the advantages of high reliability, flexible control, convenience in realizing comprehensive utilization of different types of information and the like, and is widely applied to occasions with extremely long working time and extremely high safety requirements, such as manned space flight, deep space exploration, military and civil aviation and the like.

The fault detection method of the multi-redundancy inertial navigation system in the prior art is low in detection efficiency and difficult to locate specific fault devices.

Disclosure of Invention

The technical problems to be solved by the invention are as follows: the fault detection method of the dual-redundancy inertial navigation system in the prior art is low in detection efficiency and difficult to locate specific fault devices.

The technical scheme of the invention is as follows:

a quick fault detection method for a dual-redundancy inertial navigation system is characterized by comprising the following steps: the method comprises the following steps: step 1, detecting a bottom inertial navigation fault; step 2, detecting faults in the navigation software; and 3, detecting faults in the interface control software.

The step 1 comprises the following steps:

and (3) respectively carrying out IMU inertia measurement component fault detection in the preprocessing software of the two sets of single inertial navigation systems with local references: if the detection results of the two sets of inertial navigation systems are normal, performing the step 2 and continuing the normal working process; if one set of inertial navigation system fails, switching and outputting navigation data of the other set of inertial navigation system with normal detection results; and if the two sets of inertial navigation systems are in fault, forwarding and outputting main reference data.

The detection content of the IMU inertial measurement unit fault detection comprises the following six aspects:

a) whether the temperature of the laser gyroscope in the three directions is normal or not;

b) judging whether the machine shaking signals of the laser gyroscope in the three directions are normal or not;

c) whether the starting signals of the laser gyro in the three directions are normal or not;

d) whether the sum frequency signals of the laser gyro in the three directions are normal or not;

e) whether the temperature of the quartz flexible accelerometer in the three directions is normal or not;

f) and whether the original pulse signals of the quartz flexible accelerometer in the three directions are normal or not.

The step 2 comprises the following steps:

in the local reference, the IMU inertia measurement component of each set of single inertial navigation system outputs the original pulse values of the gyroscope and the accelerometer to the corresponding navigation software, and the navigation result is detected in the navigation software: if the detection results of the two sets of inertial navigation systems are normal, performing the step 3, and continuing the normal working process; if one set of inertial navigation system fails, switching and outputting navigation data of the other set of inertial navigation system with normal detection results; and if the two sets of inertial navigation systems are in fault, forwarding and outputting main reference data.

The detection content of the navigation result detection comprises the following eight aspects:

a) whether an INT0 interrupt times out;

b) whether FLASH reading and writing are abnormal or not;

c) whether the initialization binding fails or not;

d) whether hardware initialization fails;

e) whether the accelerations in the three directions are out of tolerance:

judging the upper limit value of the acceleration output by the quartz flexible accelerometer in the x direction, the y direction and the z direction according to the following formula, and if the acceleration is out of tolerance, carrying out system failure:

$\left\{\begin{array}{c}\left|\mathrm{Ax}\right|<5g\\ \left|\mathrm{Ay}\right|<6g\\ \left|\mathrm{Az}\right|<5g\end{array}\right.$

in the formula,

ax, Ay and Az are quartz flexible accelerometers outputting acceleration in x, y and z directions in sequence;

g is the acceleration of gravity;

f) whether the angular velocities in the three directions are out of tolerance:

judging the upper limit value of the angular speed output by the laser gyro in the x direction, the y direction and the z direction according to the following formula, and if the angular speed is out of tolerance, then the system fails:

in the formula,

gx, Gy and Gz are the output angular velocities of the laser gyroscope in the x, y and z directions in sequence;

g) whether the north and east speed of the navigation solution is out of tolerance:

judging the inertial navigation output speed by adopting the following formula, and if the speed is out of tolerance, carrying out system fault:

$\sqrt{{\left(V_{\mathrm{out}}^N\right)}^2+{\left(V_{\mathrm{out}}^E\right)}^2}<V_{\max }$

in the formula,

a northbound speed resolved for navigation;

east speed resolved for navigation;

V_maxthe maximum navigational speed of the warship is 30m/s in the embodiment;

h) whether three attitude angles of navigation solution are out of tolerance:

judging the inertial navigation output attitude angle by adopting the following formula, and if the attitude angle is out of tolerance, carrying out system fault:

$\left\{\begin{array}{c}\left|\mathrm{roll}\right|<{\mathrm{Angel}}_{\max }\\ \left|\mathrm{pitch}\right|<{\mathrm{Angel}}_{\max }\end{array}\right.$

in the formula,

roll is a rolling angle of navigation calculation;

pitch is a pitch angle calculated by navigation;

Angle_maxis the maximum horizontal attitude angle of the warship.

The step 3 comprises the following steps:

step 3.1. detection of inertial error estimation kafang

Establishing an observation equation:

Z＝H·X+v

wherein,

in the formula,

the initial value of X is 0;

sequentially setting the installation error angles in the x direction, the y direction and the z direction between the two sets of inertial navigations;

sequentially carrying out gyroscope drift in the directions of x, y and z between the two sets of inertial navigation;

sequentially enabling the accelerometers in the x, y and z directions between the two sets of inertial navigation to have zero offset;

z is an observation vector;

sequentially obtaining the average angular velocities of the inertial navigation system 1 in the x, y and z directions in a fault detection period;

sequentially obtaining the average angular velocities of the inertial navigation system 2 in the x, y and z directions in a fault detection period;

sequentially obtaining the average acceleration of the inertial navigation system 1 in the x, y and z directions in a fault detection period;

sequentially obtaining the average acceleration of the inertial navigation system 2 in the x, y and z directions in a fault detection period;

h is an observation coefficient matrix;

is the average angular velocity in the x direction of the inertial navigation system 1 and the inertial navigation system 2,

is the y-direction average angular velocity of the inertial navigation system 1 and the inertial navigation system 2,

；

is the z-direction average angular velocity of the inertial navigation system 1 and the inertial navigation system 2,

v is observation noise;

residual r of the k step_kAs shown in the following formula:

$r_k=Z_k-H_k{\hat{X}}_k,$

wherein, ${\hat{X}}_k={\mathrm{\&Phi;}}_{k/k-1}{\hat{X}}_{k-1}.$

in the formula,

r_kresidual errors in the k step are obtained;

predicting value in the k step;

Φ_k/k-1is a transition matrix from the (k-1) th step to the k-th step;

let A_kIs r_kAnd (4) variance, wherein a fault detection function corresponding to each observed quantity is shown as the following formula:

$\left\{\begin{array}{c}{\mathrm{\&lambda;}}_k\_\mathrm{\&omega;x}=r_k\left(1\right)A_k^{-1}\left(\mathrm{1,1}\right)r_k\left(1\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{\&omega;y}=r_k\left(2\right)A_k^{-1}\left(\mathrm{2,2}\right)r_k\left(2\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{\&omega;z}=r_k\left(3\right)A_k^{-1}\left(\mathrm{3,3}\right)r_k\left(3\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{ax}=r_k\left(4\right)A_k^{-1}\left(\mathrm{4,4}\right)r_k\left(4\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{ay}=r_k\left(5\right)A_k^{-1}\left(\mathrm{5,5}\right)r_k\left(5\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{az}=r_k\left(6\right)A_k^{-1}\left(\mathrm{6,6}\right)r_k\left(6\right)\end{array}\right.,$

setting a threshold value of a kaer square for each observed quantity, and judging that a device corresponding to the observed quantity has a fault when a fault detection function value corresponding to each observed quantity is greater than the threshold value of the kaer square;

step 3.2. detection of specific faulty devices

Calculating the absolute value of each frame of variation output by the inertia devices of the two sets of inertial navigation systems by adopting the following formula:

$\left\{\begin{array}{c}d{\mathrm{\&omega;}}_{1i}\left(t_k\right)=|{\mathrm{\&omega;}}_{1i}\left(t_k\right)-{\mathrm{\&omega;}}_{1i}(t_k-1)|\\ {\mathrm{df}}_{1i}\left(t_k\right)=|f_{1i}\left(t_k\right)-f_{1i}(t_k-1)|\\ d{\mathrm{\&omega;}}_{2i}\left(t_k\right)=|{\mathrm{\&omega;}}_{2i}\left(t_k\right)-{\mathrm{\&omega;}}_{2i}(t_k-1)|\\ {\mathrm{df}}_{2i}\left(t_k\right)=|f_{2i}\left(t_k\right)-f_{2i}(t_k-1)|\end{array}\right.,i=x,y,z$

in the formula,

dω_1i(t_k) The angular velocity variation of the current frame of the i-direction inertial navigation system 1 is obtained;

df_1i(t_k) The current frame acceleration variation of the inertial navigation system 1 in the i direction is obtained;

dω_2i(t_k) Is the direction of inertia of iLeading the system 2 to obtain the angular speed variation of the current frame;

df_2i(t_k) The current frame acceleration variation of the inertial navigation system 2 in the i direction is obtained;

ω_1i(t_k)、ω_1i(t_k-1) sequentially outputting angular velocity values of a current frame and a previous frame of the inertial navigation system 1 in the i direction;

f_1i(t_k)、f_1i(t_k-1) sequentially outputting the acceleration of a current frame and an acceleration of a previous frame of the inertial navigation system 1 in the direction i;

ω_2i(t_k)、ω_2i(t_k-1) sequentially outputting angular velocity values of a current frame and a previous frame of the i-direction inertial navigation system 2;

f_2i(t_k)、f_2i(t_k-1) sequentially outputting the acceleration of a current frame and an acceleration of a previous frame of the inertial navigation system 2 in the direction i;

the maximum value of the absolute value of each frame of variation in a fault detection period is recorded as d omega_{1i_MAX}、df_{1i_MAX}、dω_{2i_MAX}、df_{2i_MAX}I ═ x, y, z; and subtracting the maximum value of the absolute value of the variation in the same direction of the inertial navigation system 1 and the inertial navigation system 2, namely:

$\left\{\begin{array}{c}{\mathrm{d\&omega;}}_i={\mathrm{d\&omega;}}_{\mathrm{li}\_\mathrm{MAX}}-{\mathrm{d\&omega;}}_{2i\_\mathrm{MAX}}\\ {\mathrm{df}}_i={\mathrm{df}}_{1i\_\mathrm{MAX}}-{\mathrm{df}}_{2i\_\mathrm{MAX}}\end{array}\right.$

if d ω_iOr df_iIf the value is greater than a given positive threshold value T _ positive, judging that the device of the inertial navigation system 1 is in fault; if d ω_iOr df_iIf the fault is less than the given negative threshold value T _ negative, judging that the device of the inertial navigation system 2 is in fault; otherwise, judging that the inertia devices have no fault.

The invention has the beneficial effects that:

(1) the quick fault detection method for the dual-redundancy inertial navigation system can detect the inertial devices and key signals of two sets of inertial navigation systems in the local reference in real time, has high detection efficiency and provides guarantee for normal work of the local reference;

(2) the quick fault detection method for the dual-redundancy inertial navigation system can accurately position a fault device;

(3) the quick fault detection method for the dual-redundancy inertial navigation system can quickly and accurately detect various typical inertial navigation faults and automatically switch and output navigation data of normal inertial navigation.

Drawings

Fig. 1 is a flow chart of local fiducial detection.

Detailed Description

The following describes a method for detecting a fast failure of a dual redundant inertial navigation system in detail with reference to the accompanying drawings and embodiments.

The invention discloses a quick fault detection method for a dual-redundancy inertial navigation system, which comprises the following steps of: firstly, detecting bottom layer faults in the preprocessing software of the inertia measurement components of two sets of inertial navigation systems; then, detecting the navigation result in the navigation software of the two sets of inertial navigation systems; and finally, detecting typical faults such as hop count, slow drift and the like by comparing the angular velocity and the acceleration of the two sets of inertial navigation systems in interface control software.

Step 1, bottom inertial navigation fault detection

And (3) respectively carrying out IMU inertia measurement component fault detection in the preprocessing software of the two sets of single inertial navigation systems with local references: if the detection results of the two sets of inertial navigation systems are normal, performing the step 2 and continuing the normal working process; if one set of inertial navigation system fails, switching and outputting navigation data of the other set of inertial navigation system with normal detection results; and if the two sets of inertial navigation systems are in fault, forwarding and outputting main reference data.

The detection content of the fault detection of the IMU inertial measurement unit comprises the following six aspects, and the specific detection method is common knowledge of the technical personnel in the field:

a) whether the temperature of the laser gyroscope in the three directions is normal or not;

b) judging whether the machine shaking signals of the laser gyroscope in the three directions are normal or not;

c) whether the starting signals of the laser gyro in the three directions are normal or not;

d) whether the sum frequency signals of the laser gyro in the three directions are normal or not;

e) whether the temperature of the quartz flexible accelerometer in the three directions is normal or not;

f) and whether the original pulse signals of the quartz flexible accelerometer in the three directions are normal or not.

Step 2, detecting faults in navigation software

In the local reference, the IMU inertia measurement component of each set of single inertial navigation system outputs the original pulse values of the gyroscope and the accelerometer to the corresponding navigation software, the original pulse values are subjected to calibration compensation and dynamic error compensation to obtain angular velocity and acceleration, navigation calculation is completed, and the navigation result is detected in the navigation software: if the detection results of the two sets of inertial navigation systems are normal, performing the step 3, and continuing the normal working process; if one set of inertial navigation system fails, switching and outputting navigation data of the other set of inertial navigation system with normal detection results; and if the two sets of inertial navigation systems are in fault, forwarding and outputting main reference data.

The detection content of the navigation result detection comprises the following eight aspects:

a) whether INT0 interrupt times out

INT0 interruption is navigation period interruption, and the interruption is overtime if the interruption is not interrupted after the interruption exceeds the preset navigation period INT0, which indicates system failure.

b) Whether FLASH reading and writing is abnormal or not

Reading FLASH parameters when a system is started and initialized, and if the FLASH reading is abnormal, indicating that the system is in fault; and writing FLASH parameters in the system calibration process, and if the FLASH writing is abnormal, indicating that the system has a writing fault at this time, but the system has no fault. The judgment criteria of the FLASH read exception and the write exception are common knowledge of the skilled in the art.

c) Whether or not the initial binding failed

And (3) judging the validity of external information during initial binding: if the foreign information is invalid, the initial binding fails. The interface control software needs to report the state to the missile launching coordination management machine, but the system has no fault, and the system still waits for effective information to be bound until the binding is successful.

d) Whether hardware initialization failed

And judging the hardware initialization state when the system starts initialization, and representing a system fault if the hardware initialization fails.

e) Whether the accelerations in three directions are out of tolerance

Judging the upper limit value of the acceleration output by the quartz flexible accelerometer in the x direction, the y direction and the z direction according to the following formula, and if the acceleration is out of tolerance, carrying out system failure:

$\left\{\begin{array}{c}\left|\mathrm{Ax}\right|<5g\\ \left|\mathrm{Ay}\right|<6g\\ \left|\mathrm{Az}\right|<5g\end{array}\right.$

in the formula,

ax, Ay and Az are quartz flexible accelerometers outputting acceleration in x, y and z directions in sequence;

g is the acceleration of gravity.

f) Whether the angular velocities in three directions are out of tolerance

Judging the upper limit value of the angular speed output by the laser gyro in the x direction, the y direction and the z direction according to the following formula, and if the angular speed is out of tolerance, then the system fails:

in the formula,

and Gx, Gy and Gz are the output angular speeds of the laser gyroscope in the x, y and z directions in sequence.

g) Whether the north and east speeds of the navigation solution are out of tolerance

Judging the inertial navigation output speed according to the navigation speed of the warship is an effective fault detection method, judging the inertial navigation output speed by adopting the following formula, and if the speed is out of tolerance, judging the system fault:

$\sqrt{{\left(V_{\mathrm{out}}^N\right)}^2+{\left(V_{\mathrm{out}}^E\right)}^2}<V_{\max }$

in the formula,

a northbound speed resolved for navigation;

east speed resolved for navigation;

V_maxthe maximum navigational speed of the warship is army in this embodimentThe maximum navigational speed of the ship is 30 m/s.

h) Whether three attitude angles of navigation calculation are out of tolerance

Judging the inertial navigation output attitude according to the horizontal attitude angle of the warship is an effective fault detection method, judging the inertial navigation output attitude angle by adopting the following formula, and if the attitude angle is out of tolerance, judging the system fault:

$\left\{\begin{array}{c}\left|\mathrm{roll}\right|<{\mathrm{Angle}}_{\max }\\ \left|\mathrm{pitch}\right|<{\mathrm{Angle}}_{\max }\end{array}\right.$

in the formula,

roll is a rolling angle of navigation calculation;

pitch is a pitch angle calculated by navigation;

Angle_maxthe maximum horizontal attitude angle of the warship is 50 in this embodiment. .

Step 3, detecting the fault in the interface control software

If no fault is detected in the step 1 and the step 2, sending the angular velocity and the acceleration of the two sets of inertial navigation systems to interface control software, ensuring the synchronism of the two sets of inertial navigation systems by a time system on the ship, and completing double-inertial navigation mutual judgment fault detection in the interface control software.

Step 3.1. detection of inertial error estimation kafang

And performing inertial error estimation kaider detection by adopting a Kalman filtering technology.

Establishing an observation equation:

Z=H·X+v

wherein, $\left\{\begin{array}{c}X=\left[\begin{array}{ccccccccc}{\stackrel{\&CenterDot;}{\mathrm{\&phi;}}}_{\mathrm{ax}}& {\stackrel{\&CenterDot;}{\mathrm{\&phi;}}}_{\mathrm{ay}}& {\stackrel{\&CenterDot;}{\mathrm{\&phi;}}}_{\mathrm{az}}& {\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}}_x& {\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}}_y& {\stackrel{\&CenterDot;}{\mathrm{\&epsiv;}}}_z& {\stackrel{\&CenterDot;}{\mathrm{\&Delta;}}}_x& {\stackrel{\&CenterDot;}{\mathrm{\&Delta;}}}_y& {\stackrel{\&CenterDot;}{\mathrm{\&Delta;}}}_z\end{array}\right]\\ H\begin{array}{c}\end{array}=\left[\begin{array}{ccccccccc}0& -{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_z& {\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_y& 1& 0& 0& 0& 0& 0\\ {\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_z& 0& -{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_x& 0& 1& 0& 0& 0& 0\\ -{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_y& {\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_x& 0& 0& 0& 1& 0& 0& 0\\ 0& -{\stackrel{\&OverBar;}{f}}_z& {\stackrel{\&OverBar;}{f}}_y& 0& 0& 0& 1& 0& 0\\ {\stackrel{\&OverBar;}{f}}_z& 0& -{\stackrel{\&OverBar;}{f}}_x& 0& 0& 0& 0& 1& 0\\ -{\stackrel{\&OverBar;}{f}}_y& {\stackrel{\&OverBar;}{f}}_x& 0& 0& 0& 0& 0& 0& 1\end{array}\right]\\ Z=\left[\begin{array}{c}{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_{x1}\left(t_k\right)-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_{x2}\left(t_k\right)\\ {\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_{y1}\left(t_k\right)-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_{y2}\left(t_k\right)\\ {\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_{z1}\left(t_k\right)-{\stackrel{\&OverBar;}{\mathrm{\&omega;}}}_{z2}\left(t_k\right)\\ {\stackrel{\&OverBar;}{f}}_{x1}\left(t_k\right)-{\stackrel{\&OverBar;}{f}}_{x2}\left(t_k\right)\\ {\stackrel{\&OverBar;}{f}}_{y1}\left(t_k\right)-{\stackrel{\&OverBar;}{f}}_{y2}\left(t_k\right)\\ {\stackrel{\&OverBar;}{f}}_{z1}\left(t_k\right)-{\stackrel{\&OverBar;}{f}}_{z2}\left(t_k\right)\end{array}\right]\end{array}\right.$

in the formula,

the initial value of X is 0;

sequentially setting the installation error angles in the x direction, the y direction and the z direction between the two sets of inertial navigations;

sequentially carrying out gyroscope drift in the directions of x, y and z between the two sets of inertial navigation;

sequentially enabling the accelerometers in the x, y and z directions between the two sets of inertial navigation to have zero offset;

z is an observation vector;

sequentially obtaining the average angular velocities of the inertial navigation system 1 in the x, y and z directions in a fault detection period;

sequentially obtaining the average angular velocities of the inertial navigation system 2 in the x, y and z directions in a fault detection period;

sequentially obtaining the average acceleration of the inertial navigation system 1 in the x, y and z directions in a fault detection period;

sequentially obtaining the average acceleration of the inertial navigation system 2 in the x, y and z directions in a fault detection period;

h is an observation coefficient matrix;

is the average angular velocity in the x direction of the inertial navigation system 1 and the inertial navigation system 2,

is the y-direction average angular velocity of the inertial navigation system 1 and the inertial navigation system 2,

;

is the z-direction average angular velocity of the inertial navigation system 1 and the inertial navigation system 2,

v is the observation noise.

Residual r of the k step_kAs shown in the following formula:

$r_k=Z_k-H_k{\hat{X}}_k,$

wherein, ${\hat{X}}_k={\mathrm{\&Phi;}}_{k/k-1}{\hat{X}}_{k-1}.$

in the formula,

r_kresidual errors in the k step are obtained;

predicting value in the k step;

Φ_k/k-1is the transition matrix from the (k-1) th step to the k-th step.

Let A_kIs r_kAnd (4) variance, wherein a fault detection function corresponding to each observed quantity is shown as the following formula:

$\left\{\begin{array}{c}{\mathrm{\&lambda;}}_k\_\mathrm{\&omega;x}=r_k\left(1\right)A_k^{-1}\left(\mathrm{1,1}\right)r_k\left(1\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{\&omega;y}=r_k\left(2\right)A_k^{-1}\left(\mathrm{2,2}\right)r_k\left(2\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{\&omega;z}=r_k\left(3\right)A_k^{-1}\left(\mathrm{3,3}\right)r_k\left(3\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{ax}=r_k\left(4\right)A_k^{-1}\left(\mathrm{4,4}\right)r_k\left(4\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{ay}=r_k\left(5\right)A_k^{-1}\left(\mathrm{5,5}\right)r_k\left(5\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{az}=r_k\left(6\right)A_k^{-1}\left(\mathrm{6,6}\right)r_k\left(6\right)\end{array}\right.,$

all the fault detection functions corresponding to the observed quantities are subjected to chi²And (4) distribution. When a certain device of the system has a fault and the output of the inertia device is abnormal, Z_kNumerical value of and

will have a large difference in value of (c), which in turn affects r_kYielding a larger value of the chi-square. Such as a device failure to jump, which will result in r_kJump is also generated, so that the corresponding click square value is increased, therefore, the fault can be detected by setting a click square threshold value, and the judgment criterion is as follows: and respectively setting a threshold value of a kaer square for each observed quantity, and when the fault detection function value corresponding to each observed quantity is greater than the threshold value of the kaer square, judging that the device corresponding to the observed quantity has a fault. Such as lambda_kIf omega x is larger than the kaider threshold value, the observed quantity is judged

The corresponding device fails, i.e. the device corresponding to the average angular velocity in the x-direction fails.

And (3) estimating a chi-square by adopting inertial errors, detecting to position that the inertial device in a certain direction fails, and then positioning the failed device in the inertial navigation system 1 or the inertial system 2 by the step 3.2, namely positioning the specific failed device.

Step 3.2. detection of specific faulty devices

Calculating the absolute value of each frame of variation output by the inertia devices of the two sets of inertial navigation systems by adopting the following formula:

$\left\{\begin{array}{c}{\mathrm{d\&omega;}}_{1i}\left(t_k\right)=|{\mathrm{\&omega;}}_{\mathrm{li}}\left(t_k\right)-{\mathrm{\&omega;}}_{\mathrm{li}}(t_k-1)|\\ {\mathrm{df}}_{1i}\left(t_k\right)=|f_{\mathrm{li}}\left(t_k\right)-f_{\mathrm{li}}(t_k-1)|\\ {\mathrm{d\&omega;}}_{2i}\left(t_k\right)=|{\mathrm{\&omega;}}_{2i}\left(t_k\right)-{\mathrm{\&omega;}}_{2i}(t_k-1)|\\ {\mathrm{df}}_{2i}\left(t_k\right)=|f_{2i}\left(t_k\right)-f_{2i}(t_k-1)|\end{array}\right.,(i=x,y,z)$

in the formula,

dω_1i(t_k) The angular velocity variation of the current frame of the i-direction inertial navigation system 1 is obtained;

df_1i(t_k) The current frame acceleration variation of the inertial navigation system 1 in the i direction is obtained;

dω_2i(t_k) The angular velocity variation of the current frame of the i-direction inertial navigation system 2;

df_2i(t_k) The current frame acceleration variation of the inertial navigation system 2 in the i direction is obtained;

ω_1i(t_k)、ω_1i(t_k-1) sequentially outputting angular velocity values of a current frame and a previous frame of the inertial navigation system 1 in the i direction;

f_1i(t_k)、f_1i(t_k-1) sequentially outputting the acceleration of a current frame and an acceleration of a previous frame of the inertial navigation system 1 in the direction i;

ω_2i(t_k)、ω_2i(t_k-1) sequentially outputting angular velocity values of a current frame and a previous frame of the i-direction inertial navigation system 2;

f_2i(t_k)、f_2i(t_kand-1) sequentially outputting the acceleration values of the current frame and the previous frame of the i-direction inertial navigation system 2.

In this embodiment, each fault detection period is 25ms, one frame every 5 ms.

The maximum value of the absolute value of each frame of variation in a fault detection period is recorded as d omega_{li_MAX}、df_{1i_MAX}、dω_{2i_MAX}、df_{2i_MAX}I = x, y, z; and subtracting the maximum value of the absolute value of the variation in the same direction of the inertial navigation system 1 and the inertial navigation system 2, namely:

$\left\{\begin{array}{c}{\mathrm{d\&omega;}}_i={\mathrm{d\&omega;}}_{1i\_\mathrm{MAX}}-{\mathrm{d\&omega;}}_{2i\_\mathrm{MAX}}\\ {\mathrm{df}}_i={\mathrm{df}}_{1i\_\mathrm{MAX}}-{\mathrm{df}}_{2i\_\mathrm{MAX}}\end{array}\right.,$

if d ω_iOr df_iIf the value is greater than a given positive threshold value T _ positive, judging that the device of the inertial navigation system 1 is in fault; if d ω_iOr df_iIf the fault is less than the given negative threshold value T _ negative, judging that the device of the inertial navigation system 2 is in fault; otherwise, judging that the inertia devices have no fault.

The positive threshold T _ positive and the negative threshold T _ negative are set according to actual conditions, which is common knowledge of those skilled in the art.

Claims (6)

1. A quick fault detection method for a dual-redundancy inertial navigation system is characterized by comprising the following steps: the method comprises the following steps: step 1, detecting a bottom inertial navigation fault; step 2, detecting faults in the navigation software; and 3, detecting faults in the interface control software.

2. The dual-redundancy inertial navigation system rapid fault detection method according to claim 1, characterized in that: the step 1 comprises the following steps:

and (3) respectively carrying out IMU inertia measurement component fault detection in the preprocessing software of the two sets of single inertial navigation systems with local references: if the detection results of the two sets of inertial navigation systems are normal, performing the step 2 and continuing the normal working process; if one set of inertial navigation system fails, switching and outputting navigation data of the other set of inertial navigation system with normal detection results; and if the two sets of inertial navigation systems are in fault, forwarding and outputting main reference data.

3. The dual-redundancy inertial navigation system rapid fault detection method according to claim 2, characterized in that: in step 1, the detection content of fault detection of the IMU inertial measurement unit includes the following six aspects:

a) whether the temperature of the laser gyroscope in the three directions is normal or not;

b) judging whether the machine shaking signals of the laser gyroscope in the three directions are normal or not;

c) whether the starting signals of the laser gyro in the three directions are normal or not;

d) whether the sum frequency signals of the laser gyro in the three directions are normal or not;

e) whether the temperature of the quartz flexible accelerometer in the three directions is normal or not;

f) and whether the original pulse signals of the quartz flexible accelerometer in the three directions are normal or not.

4. The method for detecting the fast failure of the dual-redundancy inertial navigation system according to claim 1 or 2, wherein: the step 2 comprises the following steps:

in the local reference, the IMU inertia measurement component of each set of single inertial navigation system outputs the original pulse values of the gyroscope and the accelerometer to the corresponding navigation software, and the navigation result is detected in the navigation software: if the detection results of the two sets of inertial navigation systems are normal, performing the step 3, and continuing the normal working process; if one set of inertial navigation system fails, switching and outputting navigation data of the other set of inertial navigation system with normal detection results; and if the two sets of inertial navigation systems are in fault, forwarding and outputting main reference data.

5. The dual-redundancy inertial navigation system rapid fault detection method according to claim 4, characterized in that: in step 2, the detection content of the navigation result detection includes the following eight aspects:

a) whether an INT0 interrupt times out;

b) whether FLASH reading and writing are abnormal or not;

c) whether the initialization binding fails or not;

d) whether hardware initialization fails;

e) whether the accelerations in the three directions are out of tolerance:

judging the upper limit value of the acceleration output by the quartz flexible accelerometer in the x direction, the y direction and the z direction according to the following formula, and if the acceleration is out of tolerance, carrying out system failure:

$\left\{\begin{array}{c}\left|\mathrm{Ax}\right|<5g\\ \left|\mathrm{Ay}\right|<6g\\ \left|\mathrm{Az}\right|<5g\end{array}\right.$

in the formula,

ax, Ay and Az are quartz flexible accelerometers outputting acceleration in x, y and z directions in sequence;

g is the acceleration of gravity;

f) whether the angular velocities in the three directions are out of tolerance:

judging the upper limit value of the angular speed output by the laser gyro in the x direction, the y direction and the z direction according to the following formula, and if the angular speed is out of tolerance, then the system fails:

in the formula,

gx, Gy and Gz are the output angular velocities of the laser gyroscope in the x, y and z directions in sequence;

g) whether the north and east speed of the navigation solution is out of tolerance:

judging the inertial navigation output speed by adopting the following formula, and if the speed is out of tolerance, carrying out system fault:

$\sqrt{{\left(V_{\mathrm{out}}^N\right)}^2+{\left(V_{\mathrm{out}}^E\right)}^2}<V_{\max }$

in the formula,

a northbound speed resolved for navigation;

east speed resolved for navigation;

V_maxthe maximum navigational speed of the warship is 30m/s in the embodiment;

h) whether three attitude angles of navigation solution are out of tolerance:

judging the inertial navigation output attitude angle by adopting the following formula, and if the attitude angle is out of tolerance, carrying out system fault:

$\left\{\begin{array}{c}\left|\mathrm{roll}\right|<\mathrm{Angl}{e}_{\max }\\ \left|\mathrm{pitch}\right|<{\mathrm{Angle}}_{\max }\end{array}\right.$

in the formula,

roll is a rolling angle of navigation calculation;

pitch is a pitch angle calculated by navigation;

Angle_maxis the maximum horizontal attitude angle of the warship.

6. The dual-redundancy inertial navigation system rapid fault detection method according to claim 4, characterized in that: the step 3 comprises the following steps:

step 3.1. detection of inertial error estimation kafang

Establishing an observation equation:

Z＝H.X+v

wherein,

in the formula,

the initial value of X is 0;

sequentially setting the installation error angles in the x direction, the y direction and the z direction between the two sets of inertial navigations;

sequentially carrying out gyroscope drift in the directions of x, y and z between the two sets of inertial navigation;

sequentially enabling the accelerometers in the x, y and z directions between the two sets of inertial navigation to have zero offset;

z is an observation vector;

sequentially obtaining the average angular velocities of the inertial navigation system 1 in the x, y and z directions in a fault detection period;

sequentially obtaining the average angular velocities of the inertial navigation system 2 in the x, y and z directions in a fault detection period;

sequentially obtaining the average acceleration of the inertial navigation system 1 in the x, y and z directions in a fault detection period;

sequentially obtaining the average acceleration of the inertial navigation system 2 in the x, y and z directions in a fault detection period;

h is an observation coefficient matrix;

is the average angular velocity in the x direction of the inertial navigation system 1 and the inertial navigation system 2,

is the y-direction average angular velocity of the inertial navigation system 1 and the inertial navigation system 2,

；

is the z-direction average angular velocity of the inertial navigation system 1 and the inertial navigation system 2,

v is observation noise;

residual r of the k step_kAs shown in the following formula:

$r_k=Z_k-H_k{\hat{X}}_k,$

wherein, ${\hat{X}}_k={\mathrm{\&Phi;}}_{k/k-1}{\hat{X}}_{k-1}.$

in the formula,

r_kresidual errors in the k step are obtained;

predicting value in the k step;

Φ_k/k-1is a transition matrix from the (k-1) th step to the k-th step;

let A_kIs r_kAnd (4) variance, wherein a fault detection function corresponding to each observed quantity is shown as the following formula:

$\left\{\begin{array}{c}{\mathrm{\&lambda;}}_k\_\mathrm{\&omega;x}=r_k\left(1\right)A_k^{-1}\left(\mathrm{1,1}\right)r_k\left(1\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{\&omega;y}=r_k\left(2\right)A_k^{-1}\left(\mathrm{2,2}\right)r_k\left(2\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{\&omega;z}=r_k\left(3\right)A_k^{-1}\left(\mathrm{3,3}\right)r_k\left(3\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{ax}=r_k\left(4\right)A_k^{-1}\left(\mathrm{4,4}\right)r_k\left(4\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{ay}=r_k\left(5\right)A_k^{-1}\left(\mathrm{5,5}\right)r_k\left(5\right)\\ {\mathrm{\&lambda;}}_k\_\mathrm{az}=r_k\left(6\right)A_k^{-1}\left(\mathrm{6,6}\right)r_k\left(6\right)\end{array}\right.,$

setting a threshold value of a kaer square for each observed quantity, and judging that a device corresponding to the observed quantity has a fault when a fault detection function value corresponding to each observed quantity is greater than the threshold value of the kaer square;

step 3.2. detection of specific faulty devices

Calculating the absolute value of each frame of variation output by the inertia devices of the two sets of inertial navigation systems by adopting the following formula:

$\left\{\begin{array}{c}{\mathrm{d\&omega;}}_{1i}\left(t_k\right)=|{\mathrm{\&omega;}}_{1i}\left(t_k\right)-{\mathrm{\&omega;}}_{1i}(t_k-1)|\\ {\mathrm{df}}_{1i}\left(t_k\right)=|f_{1i}\left(t_k\right)-f_{1i}(t_k-1)|\\ {\mathrm{d\&omega;}}_{2i}\left(t_k\right)=|{\mathrm{\&omega;}}_{2i}\left(t_k\right)-{\mathrm{\&omega;}}_{2i}(t_k-1)|\\ {\mathrm{df}}_{2i}\left(t_k\right)=|f_{2i}\left(t_k\right)-f_{2i}(t_k-1)|\end{array}\right.,i=x,y,z$

in the formula,

dω_1i(t_k) The angular velocity variation of the current frame of the i-direction inertial navigation system 1 is obtained;

df_1i(t_k) The current frame acceleration variation of the inertial navigation system 1 in the i direction is obtained;

dω_2i(t_k) The angular velocity variation of the current frame of the i-direction inertial navigation system 2;

df_2i(t_k) The current frame acceleration variation of the inertial navigation system 2 in the i direction is obtained;

ω_1i(t_k)、ω_1i(t_k-1) sequentially outputting angular velocity values of a current frame and a previous frame of the inertial navigation system 1 in the i direction;

f_1i(t_k)、f_1i(t_k-1) sequentially outputting the acceleration of a current frame and an acceleration of a previous frame of the inertial navigation system 1 in the direction i;

ω_2i(t_k)、ω_2i(t_k-1) sequentially outputting angular velocity values of a current frame and a previous frame of the i-direction inertial navigation system 2;

f_2i(t_k)、f_2i(t_k-1) sequentially outputting the acceleration of a current frame and an acceleration of a previous frame of the inertial navigation system 2 in the direction i;

the maximum value of the absolute value of each frame of variation in a fault detection period is recorded as d omega_{1i_MAX}、df_{1i_MAX}、dω_{2i_MAX}、df_{2i_MAX}I ═ x, y, z; and subtracting the maximum value of the absolute value of the variation in the same direction of the inertial navigation system 1 and the inertial navigation system 2, namely:

$\left\{\begin{array}{c}{\mathrm{d\&omega;}}_i={\mathrm{d\&omega;}}_{1i\_\mathrm{MAX}}-{\mathrm{d\&omega;}}_{2i\_\mathrm{MAX}}\\ {\mathrm{df}}_i={\mathrm{df}}_{1i\_\mathrm{MAX}}-{\mathrm{df}}_{2i\_\mathrm{MAX}}\end{array}\right.$

if d ω_iOr df_iIf the value is greater than a given positive threshold value T _ positive, judging that the device of the inertial navigation system 1 is in fault; if d ω_iOr df_iIf the fault is less than the given negative threshold value T _ negative, judging that the device of the inertial navigation system 2 is in fault; otherwise, judging that the inertia devices have no fault.

Images