CN103499354A - Neyman-Pearson criterion-based zero speed detection method - Google Patents

Abstract

The invention discloses a Neyman-Pearson criterion-based zero speed detection method, which comprises the following steps of receiving measured information output by sensors in real time during footstep motions in a single-soldier navigation system by using a handheld computer; determining a window function N according to system sampling frequency and data transmission speed; converting a zero speed detection problem into a modeled mathematical problem by utilizing a double-hypothesis testing theory, and obtaining a Neyman-Pearson criterion-based zero detection inequality; determining mathematical models of output signals of miniature inertia measurement unit sensors and received signals of the handheld computer; calculating a joint probability density function of the output signals of the miniature inertia measurement unit sensors; replacing unknown elements in the inequality with maximum likelihood estimation values of unknown signal elements to obtain an extensive probability likelihood ratio inequality; and substituting the output data of a miniature inertia measurement unit into the extensive probability likelihood ratio inequality, and further detecting a zero speed state. According to the method, problems are mathematized and modeled, and the detection accuracy is improved.

Description

A kind of zero-speed detection method based on Nei Man-Pearson criterion

Technical field

The invention belongs to individual soldier's autonomous navigation system zero-speed detection technique field, relate in particular to a kind of zero-speed detection method based on Nei Man-Pearson criterion.

Background technology

When the individual soldier who resolves based on mini inertia measurement unit (Micro Inertial Measurement Unit) inertia navigates freedom positioning device work, micro mechanical system (Micro-Electro-Mechanical System) inertial device error is dispersed seriously, resolved individual soldier's dead reckoning result verification that algorithm obtains in navigation stage by inertia, if the Micromachined Inertial Devices error can not obtain effective compensation, site error can be dispersed with three cubed trend of time, system is lost navigation feature the most at last, therefore, micro mechanical system inertia resolves algorithm application and is to design effective error correction algorithms in the maximum difficult point of individual soldier's navigational system, it is a kind of effective Error Compensation Algorithm that zero-speed is proofreaied and correct, while utilizing normal person's walking, all there are the static and two kinds of mode of motion of gait motion (two kinds of exercise durations are all about 0.5 second) of gait in each pin, individual soldier's navigational system based on micro inertial measurement unit is connected firmly on pedestrian's footwear, be referred to as " navigation footwear ", when step contacts to earth, inertia resolves the speed theoretical value obtained and should be zero, but resolve and can access speed component by inertia, the velocity error that this speed component is totally caused in time by micro-mechanical inertia measuring element error, this speed component is input to zero-speed correction error compensator to be proofreaied and correct inertia device measurement result and navigation output,

Yet, individual soldier's navigational system that existing employing zero-speed is proofreaied and correct as inertia resolution error correction algorithm mostly exists zero-speed to detect inaccurate problem, it is to trigger the prerequisite that zero-speed is proofreaied and correct that zero-speed detects, zero-speed detection scheme in the past is mainly to utilize the output of three axis accelerometer or three axle gyros to be less than certain threshold value within the time of setting, be judged to be the zero-speed interval, choosing of its threshold value lacks theoretical research.In actual measurement, zero-speed interval when detection method in the past can only detect normal walking, during running, the detection in zero-speed interval is relatively difficult, and existing zero-speed detection method is all under ad hoc fashion, be not to be applicable to all individual soldier's motion states, as acceleration of motion variance detection scheme and be not suitable for uniformly accelrated rectilinear motion, the acceleration amplitude detection scheme cannot be for detection of single pin the situation around a certain axis rotation.

The poor stability of the method that existing zero-speed detects on the whole, cause zero-speed to be proofreaied and correct after navigation accuracy still lower, be difficult to meet individual soldier's accurately requirement reliably of navigating.

Summary of the invention

The purpose of the embodiment of the present invention is to provide a kind of zero-speed detection method based on Nei Man-Pearson criterion, be intended to solve poor stability, accuracy that existing zero-speed detection method exists low, after causing zero-speed to be proofreaied and correct, navigation accuracy is lower, is difficult to meet the problem that the individual soldier navigates and accurately requires reliably.

The embodiment of the present invention is achieved in that a kind of zero-speed detection method based on Nei Man-Pearson criterion, should the zero-speed detection method based on Nei Man-Pearson criterion comprise the following steps:

Step 1, hand-held palm PC receives and stores in real time the measurement information of step mini inertia measurement unit output in individual soldier's autonomous navigation system;

Step 2: data transfer rate between mini inertia measurement unit and palm PC in mini inertia measurement unit length rest time and step 1 measuring process before using according to the sample frequency of each sensor in individual soldier's autonomous navigation system, system, obtain the concrete value of window function N in this zero-speed testing process, N is integer;

Step 3: the mini inertia measurement unit output data and the definite window function N of step 2 that utilize step 1 to collect, the zero-speed test problems is converted into to modeled pair of Hypothesis Testing Problem, testing result is hypothesis mini inertia measurement unit motion H ₀, the static H of mini inertia measurement unit ₁one of, utilize Nei Man-Pearson came theorem to obtain the mathematical model that zero-speed detects judgement: known spurious alarm probability P _fAduring=α, if meet zero-speed judgement inequality, be

h ₁be true, individual soldier's autonomous navigation system zero-speed testing result is that the step mini inertia measurement unit is static; Function L (z _n) relevant with likelihood ratio, for z _neach value, meaned H ₁the hypothetical probabilities value is to H ₀the ratio of hypothetical probabilities value;

Wherein, spurious alarm probability P _fA={ H ₁∣ H ₀mean that mini inertia measurement unit is judged as the probability of stationary state while being motion state; z _n={ y _kthe n+N-1k=n measured value that is mini inertia measurement unit in a period of time; The γ value is by formula

definition; p(z _n; H ₀), p (z _n; H ₁) mean respectively the probability density function of observation data under two kinds of hypothesis;

Step 4: the signal that utilizes mini inertia measurement unit sensor output signal characteristics, palm PC to receive and the character of error disturbing signal, and formula, the mathematical model of acquisition mini inertia measurement unit inertial sensor output signal;

Step 5: utilize mathematical model and the formula of mini inertia measurement unit inertial sensor output signal definite in the data, step 4 of the output of mini inertia measurement unit in the individual soldier's autonomous navigation system motion process collected in step 1, obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Step 6: utilize the observation signal joint probability density function p (z obtained in step 5 _n; θ _i, H _i), according to formula, obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Step 7: utilize formula:

${\widehat{\mathrm{\θ}}}^0=\arg \max \left(p(z_n;{\mathrm{\θ}}_0,H_0)\right)$

${\widehat{\mathrm{\θ}}}^1=\arg \max \left(p(z_n;{\mathrm{\θ}}_1,H_1)\right)$

The maximum likelihood that obtains unknown signaling element under two kinds of assumed conditions is estimated

In formula,

at hypothesis H ₁the maximum likelihood of condition unknown element is estimated;

at hypothesis H ₀under condition, the maximum likelihood of unknown element is estimated; p(z _n; θ ₁, H ₁) be illustrated in and suppose H ₁under condition, the unknown signaling element set is θ ₁the time mini inertia measurement unit output information joint probability density function; p(z _n; θ ₀, H ₀) be illustrated in and suppose H ₀under condition, the unknown signaling element set is θ ₀the time mini inertia measurement unit output information joint probability density function; The value of unknown element set when argmax () means to make in bracket to get maximal value;

$p(z_n;{\mathrm{\θ}}_1,H_1)=\underset{k\∈{\mathrm{\Ω}}_n}{\mathrm{\Π}}p(y_k^a;{\mathrm{\θ}}_1,H_1)p(y_k^{\mathrm{\ω}};{\mathrm{\θ}}_1,H_1)$

$p(z_n,{\mathrm{\θ}}_0,H_0)=\underset{k\∈{\mathrm{\Ω}}_n}{\mathrm{\Π}}p(y_k^a;{\mathrm{\θ}}_0,H_0)p(y_k^{\mathrm{\ω}};{\mathrm{\θ}}_0,H_0);$

Step 8: the maximum likelihood that uses step 7 to try to achieve the unknown signaling element is estimated step of replacing six L _g(z _n) in the unknown signaling element set obtain not the extensive probability likelihood ratio L containing unknown element _g(z _n);

Step 9: utilize the L obtained in step 8 _g(z _n) and inequality, determine individual soldier's autonomous navigation system zero-speed testing result.

Further, in step 1, the total output information of the mini inertia measurement unit that any time k receives: for:

$y_k={\left[\begin{array}{cc}{y}_k^a& y_k^{\mathrm{\ω}}\end{array}\right]}^T;$

Wherein, T means matrix transpose operation, $y_k^{\mathrm{\ω}}={\left[\begin{array}{ccc}{y}_k^{\mathrm{\ωx}}& y_k^{\mathrm{\ωy}}& y_k^{\mathrm{\ωz}}\end{array}\right]}^T$ Angular speed information for the output of micro mechanical system three-axis gyroscope; $y_k^a={\left[\begin{array}{ccc}{y}_k^{\mathrm{ax}}& y_k^{\mathrm{ay}}& y_k^{\mathrm{az}}\end{array}\right]}^T$ Ratio force information for the output of micro mechanical system three axis accelerometer.

Further, in step 4, pass through formula:

y _k＝s _k(θ)+v _k

Obtain the mathematical model of mini inertia measurement unit inertial sensor output signal;

In formula, $s_k\left(\mathrm{\θ}\right)={\left[\begin{array}{cc}{s}_k^a\left(\mathrm{\θ}\right)& s_k^{\mathrm{\ω}}\left(\mathrm{\θ}\right)\end{array}\right]}^T,$ $v_k={\left[\begin{array}{cc}{v}_k^a& v_k^{\mathrm{\ω}}\end{array}\right]}^T,$ $s_k^a\left(\mathrm{\θ}\right)\∈{\mathrm{\Ω}}^3$ 、 $s_k^{\mathrm{\ω}}\left(\mathrm{\θ}\right)\∈{\mathrm{\Ω}}^3$ Mean respectively k mini inertia measurement unit measures constantly specific force and angular speed information, θ means the set of unknown element in needs description signal;

with

the noise set that expression is relevant to accelerometer and gyroscope respectively, T means matrix transpose operation; Suppose that noise is independent, zero-mean Gaussian distribution, covariance matrix is $C=E\left\{\begin{array}{cc}{v}_k& {v_k}^T\end{array}\right\}=\mathrm{diag}{\left[\begin{array}{cc}{{\mathrm{\σ}}_a}^2{I}_3& {{\mathrm{\σ}}_{\mathrm{\ω}}}^2{I}_3\end{array}\right]}^T,$ I wherein ₃(0 ₃) mean 3 * 3 unit (zero) battle array; Diag[] the expression diagonal matrix; E{} means to ask expectation, with

mean respectively micro-mechanical accelerometer and gyrostatic noise variance.

Further, in step 5, pass through formula:

$p(z_n,{\mathrm{\θ}}_i,H_i)=\underset{k\∈{\mathrm{\Ω}}_n}{\mathrm{\Π}}p(y_k;{\mathrm{\θ}}_i,H_i)$

$=\underset{k\∈{\mathrm{\Ω}}_n}{\mathrm{\Π}}p(y_k^a;{\mathrm{\θ}}_i,H_i)p(y_k^{\mathrm{\ω}};{\mathrm{\θ}}_i,H_i)$

Obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Wherein,

i=0,1; θ _ibe illustrated in hypothesis H _iunknown signaling element set under condition, Π means even to take advantage of; p(z _n; θ _i, H _i) be illustrated in and suppose H _iunder condition, the unknown signaling element set is θ _ithe time individual soldier autonomous navigation system output signal joint probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical accelerometer output probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical gyroscope output probability density function;

$p(y_k^a;{\mathrm{\θ}}_i,H_i)=\frac{1}{{\left(2\mathrm{\π}{\mathrm{\σ}}_a^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\σ}}_a^2}{\left|\right|y_k^a-s_k^a\left({\mathrm{\θ}}_i\right)\left|\right|}^2\}$

$p(y_k^{\mathrm{\ω}};{\mathrm{\θ}}_i,H_i)=\frac{1}{{\left(2\mathrm{\π}{\mathrm{\σ}}_{\mathrm{\ω}}^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\σ}}_{\mathrm{\ω}}^2}{\left|\right|y_k^{\mathrm{\ω}}-s_k^{\mathrm{\ω}}\left({\mathrm{\θ}}_i\right)\left|\right|}^2\}$

Wherein, exp{.} means the exponential function of e; || .|| means to ask norm; σ 2a, σ 2 ω mean respectively accelerometer and gyrostatic noise variance.

Further, in step 6, according to formula:

$L_g\left(z_n\right)=\frac{\underset{k\∈{\mathrm{\Ω}}_n}{\mathrm{\Π}}p(y_k^a;{\mathrm{\θ}}_1,H_1)p(y_k^{\mathrm{\ω}};{\mathrm{\θ}}_1,H_1)}{\underset{k\∈{\mathrm{\Ω}}_n}{\mathrm{\Π}}p(y_k^a;{\mathrm{\θ}}_0,H_0)p(y_k^{\mathrm{\ω}};{\mathrm{\θ}}_0,H_0)}$

Obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Wherein, at hypothesis H ₁conditioned signal s _k(θ) fully unknown, the set of unknown element is:

${\mathrm{\θ}}_0\≡\{z_k{\}}_{k=n}^{n+N-1};$

At hypothesis H ₀it is unknown that condition only has than the direction of force vector, and the set of unknown element is:

θ ₁≡gu _n，

In formula, u _n∈ Ω _u, Ω _u={ u ∈ R ³: ‖ u ‖=1}; G is the local gravity vector;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical accelerometer output signal probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical accelerometer output signal probability density function;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical gyroscope output signal probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical gyroscope output signal probability density function.

Further, in step 8, extensive probability likelihood ratio L _g(z _n) be expressed as;

$L_G\left(z_n\right)=\frac{p(z_n;{\widehat{\mathrm{\θ}}}^1,H_1)}{p(z_n;{\widehat{\mathrm{\θ}}}^0,H_0)}$

In formula,

be illustrated in hypothesis H ₁under condition, the unknown signaling element set is

the time mini inertia measurement unit output information joint probability density function;

be illustrated in hypothesis H ₀under condition, the unknown signaling element set is the time mini inertia measurement unit output information joint probability density function.

Further, in step 9, inequality is expressed as:

$T\left(z_n\right)=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}(\frac{1}{{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-g\frac{{\stackrel{-}{y}}_n^a}{\left|\right|{\stackrel{-}{y}}_n^a\left|\right|}\left|\right|}^2+\frac{1}{{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}\left|\right|}^2)<{\mathrm{\&gamma;}}^{\&prime;}$

If inequality is set up, individual soldier's autonomous navigation system zero-speed testing result is static-H ₁be true, otherwise the zero-speed testing result is motion-H ₀be true;

Wherein, ${\stackrel{-}{y}}_n^a=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}{y}_k^a,$ $T\left(z_n\right)=-\frac{2}{N}\ln L_G\left(z_n\right),$ γ '=-(2/N) ln (γ), ln (.) means to ask take the logarithm that e is the truth of a matter.

Zero-speed detection method based on Nei Man-Pearson criterion provided by the invention, by utilizing Nei Man-Pearson criterion, the mode of using two test of hypothesis and maximum likelihood to estimate makes the mathematicization of detection method problem, modelling.The theoretical analysis detection scheme that uses statistics to detect, overcome the shortcoming that in traditional threshold test scheme, theoretical property is poor, stability is low, under the condition that does not increase system cost, improved the precision detected; The present invention adopts the statistics etection theory analyzing and testing results such as test of hypothesis, the zero-speed that zero-speed detection method based on Nei Man-Pearson criterion is applicable in the multi-motion situation detects, and has overcome traditional zero-speed detection method low shortcoming of accuracy under the sudden change maneuver model.The inventive method is simple, and stability and reliability are high, has improved a kind of method of easy detection zero-speed, effectively raises the accuracy of zero-speed alignment technique.

The accompanying drawing explanation

Fig. 1 is the process flow diagram of the zero-speed detection method based on Nei Man-Pearson criterion that provides of the embodiment of the present invention;

Fig. 2 is pedestrian's walking (30～35s) of providing of the embodiment of the present invention and the schematic diagram of Y-axis gyro output valve while running (35～38s);

Fig. 3 is the walking that provides of the embodiment of the present invention and the interval testing result of distinct methods zero-speed schematic diagram relatively while running;

A: real zero-speed interval when walking and running; B: the zero-speed interval that when walking and running, the variance threshold values method detects; C: the zero-speed interval that the present invention is based on the zero-speed detection method detection of N-P criterion when walking and running;

Fig. 4 is that the pedestrian that the embodiment of the present invention provides jogs (5～11s) and the schematic diagram of Y-axis gyro output valve while hurrying up (11～14s);

Fig. 5 be the embodiment of the present invention provide jog and hurry up the time the interval testing result of distinct methods zero-speed schematic diagram relatively;

A: real zero-speed interval when jogging and hurrying up; B: the zero-speed interval that when jogging and hurrying up, the variance threshold values method detects; C: the zero-speed interval that the present invention is based on the zero-speed detection method detection of N-P criterion when jogging and hurrying up.

Embodiment

In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with embodiment, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

Fig. 1 shows the zero-speed detection method flow process based on Nei Man-Pearson criterion provided by the invention.For convenience of explanation, only show part related to the present invention.

The zero-speed detection method based on Nei Man-Pearson criterion of the embodiment of the present invention should the zero-speed detection method based on Nei Man-Pearson criterion comprise the following steps:

Step 1, hand-held palm PC receives and stores in real time the measurement information of step mini inertia measurement unit output in individual soldier's autonomous navigation system;

Step 2: data transfer rate between mini inertia measurement unit and palm PC in mini inertia measurement unit length rest time and step 1 measuring process before using according to the sample frequency of each sensor in individual soldier's autonomous navigation system, system, obtain the concrete value of window function N in this zero-speed testing process, N is integer;

Step 3: the mini inertia measurement unit output data and the definite window function N of step 2 that utilize step 1 to collect, the zero-speed test problems is converted into to modeled pair of Hypothesis Testing Problem, testing result is hypothesis mini inertia measurement unit motion H ₀, the static H of mini inertia measurement unit ₁one of, utilize Nei Man-Pearson came theorem to obtain the mathematical model that zero-speed detects judgement: known spurious alarm probability P _fAduring=α, if meet zero-speed judgement inequality, be

h ₁be true, individual soldier's autonomous navigation system zero-speed testing result is that the step mini inertia measurement unit is static; Function L (z _n) relevant with likelihood ratio, for z _neach value, meaned H ₁the hypothetical probabilities value is to H ₀the ratio of hypothetical probabilities value;

Wherein, spurious alarm probability P _fA={ H ₁∣ H ₀mean that mini inertia measurement unit is judged as the probability of stationary state while being motion state; z _n={ y _kthe n+N-1k=n measured value that is mini inertia measurement unit in a period of time; The γ value is by formula

definition; p(z _n; H ₀), p (z _n; H ₁) mean respectively the probability density function of observation data under two kinds of hypothesis;

Step 4: the signal that utilizes mini inertia measurement unit sensor output signal characteristics, palm PC to receive and the character of error disturbing signal, and formula, the mathematical model of acquisition mini inertia measurement unit inertial sensor output signal;

Step 5: utilize mathematical model and the formula of mini inertia measurement unit inertial sensor output signal definite in the data, step 4 of the output of mini inertia measurement unit in the individual soldier's autonomous navigation system motion process collected in step 1, obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Step 6: utilize the observation signal joint probability density function obtained in step 5

P(z _n; θ _i, H _i), according to

Formula, obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Step 7: utilize formula:

${\widehat{\mathrm{\&theta;}}}^0=\arg \max \left(p(z_n;{\mathrm{\&theta;}}_0,H_0)\right)$

${\widehat{\mathrm{\&theta;}}}^1=\arg \max \left(p(z_n;{\mathrm{\&theta;}}_1,H_1)\right)$

The maximum likelihood that obtains unknown signaling element under two kinds of assumed conditions is estimated

In formula, at hypothesis H ₁the maximum likelihood of condition unknown element is estimated;

at hypothesis H ₀under condition, the maximum likelihood of unknown element is estimated; p(z _n; θ ₁, H ₁) be illustrated in and suppose H ₁under condition, the unknown signaling element set is θ ₁the time mini inertia measurement unit output information joint probability density function; p(z _n; θ ₀, H ₀) be illustrated in and suppose H ₀under condition, the unknown signaling element set is θ ₀the time mini inertia measurement unit output information joint probability density function; The value of unknown element set when argmax () means to make in bracket to get maximal value;

$p(z_n;{\mathrm{\&theta;}}_1,H_1)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_1,H_1)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_1,H_1)$

$p(z_n,{\mathrm{\&theta;}}_0,H_0)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_0,H_0)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_0,H_0);$

Step 8: the maximum likelihood that uses step 7 to try to achieve the unknown signaling element is estimated step of replacing six L _g(z _n) in the unknown signaling element set obtain not the extensive probability likelihood ratio L containing unknown element _g(z _n);

Step 9: utilize the L obtained in step 8 _g(z _n) and inequality, determine individual soldier's autonomous navigation system zero-speed testing result.

As a prioritization scheme of the embodiment of the present invention, in step 1, the total output information of the mini inertia measurement unit that any time k receives is: $y_k={\left[\begin{array}{cc}{y}_k^a& y_k^{\mathrm{\&omega;}}\end{array}\right]}^T;$

Wherein, T means matrix transpose operation, $y_k^{\mathrm{\&omega;}}={\left[\begin{array}{ccc}{y}_k^{\mathrm{\&omega;x}}& y_k^{\mathrm{\&omega;y}}& y_k^{\mathrm{\&omega;z}}\end{array}\right]}^T$ Angular speed information for the output of micro mechanical system three-axis gyroscope; $y_k^a={\left[\begin{array}{ccc}{y}_k^{\mathrm{ax}}& y_k^{\mathrm{ay}}& y_k^{\mathrm{az}}\end{array}\right]}^T$ Ratio force information for the output of micro mechanical system three axis accelerometer.

A prioritization scheme as the embodiment of the present invention, in step 4, pass through formula:

y _k＝s _k(θ)+v _k

Obtain the mathematical model of mini inertia measurement unit inertial sensor output signal;

In formula, $s_k\left(\mathrm{\&theta;}\right)={\left[\begin{array}{cc}{s}_k^a\left(\mathrm{\&theta;}\right)& s_k^{\mathrm{\&omega;}}\left(\mathrm{\&theta;}\right)\end{array}\right]}^T,$ $v_k={\left[\begin{array}{cc}{v}_k^a& v_k^{\mathrm{\&omega;}}\end{array}\right]}^T,$ $s_k^a\left(\mathrm{\&theta;}\right)\&Element;{\mathrm{\&Omega;}}^3$ $s_k^{\mathrm{\&omega;}}\left(\mathrm{\&theta;}\right)\&Element;{\mathrm{\&Omega;}}^3$ Mean respectively k mini inertia measurement unit measures constantly specific force and angular speed, θ means the set of unknown element in needs description signal; with

the noise set that expression is relevant to accelerometer and gyroscope respectively, T means matrix transpose operation; Suppose that noise is independent, zero-mean Gaussian distribution, covariance matrix is $C=E\left\{\begin{array}{cc}{v}_k& {v_k}^T\end{array}\right\}=\mathrm{diag}{\left[\begin{array}{cc}{{\mathrm{\&sigma;}}_a}^2{I}_3& {{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}}^2{I}_3\end{array}\right]}^T,$ I wherein ₃(0 ₃) mean 3 * 3 unit (zero) battle array; Diag[] the expression diagonal matrix; E{} means to ask expectation,

with

mean respectively micro-mechanical accelerometer and gyrostatic noise variance.

A prioritization scheme as the embodiment of the present invention, in step 5, pass through formula:

$p(z_n,{\mathrm{\&theta;}}_i,H_i)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k;{\mathrm{\&theta;}}_i,H_i)$

$=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_i,H_i)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_i,H_i)$

Obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Wherein, i=0,1; θ _ibe illustrated in hypothesis H _iunknown signaling element set under condition, Π means even to take advantage of; p(z _n; θ _i, H _i) be illustrated in and suppose H _iunder condition, the unknown signaling element set is θ _ithe time individual soldier autonomous navigation system output signal joint probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical accelerometer output information probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical gyroscope output information probability density function;

$p(y_k^a;{\mathrm{\&theta;}}_i,H_i)=\frac{1}{{\left(2\mathrm{\&pi;}{\mathrm{\&sigma;}}_a^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-s_k^a\left({\mathrm{\&theta;}}_i\right)\left|\right|}^2\}$

$p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_i,H_i)=\frac{1}{{\left(2\mathrm{\&pi;}{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}-s_k^{\mathrm{\&omega;}}\left({\mathrm{\&theta;}}_i\right)\left|\right|}^2\}$

Wherein, exp{.} means the exponential function of e; || .|| means to ask norm; σ 2a, σ 2 ω mean respectively accelerometer and gyrostatic noise variance.

As a prioritization scheme of the embodiment of the present invention, in step 6, according to formula:

$L_g\left(z_n\right)=\frac{\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_1,H_1)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_1,H_1)}{\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_0,H_0)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_0,H_0)}$

Obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Wherein, at hypothesis H ₁conditioned signal s _k(θ) fully unknown, the set of unknown element is:

${\mathrm{\&theta;}}_0\&equiv;\{z_k{\}}_{k=n}^{n+N-1};$

At hypothesis H ₀it is unknown that condition only has than the direction of force vector, and the set of unknown element is:

θ ₁≡gu _n，

In formula, u _n∈ Ω _u, Ω _u={ u ∈ R ³: ‖ u ‖=1}; G is the local gravity vector;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical accelerometer output probability density function; be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical accelerometer output probability density function;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical gyroscope output probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical gyroscope output probability density function.

As a prioritization scheme of the embodiment of the present invention, in step 8, extensive probability likelihood ratio L _g(z _n) be expressed as;

$L_G\left(z_n\right)=\frac{p(z_n;{\widehat{\mathrm{\&theta;}}}^1,H_1)}{p(z_n;{\widehat{\mathrm{\&theta;}}}^0,H_0)}$

In formula,

be illustrated in hypothesis H ₁under condition, the unknown signaling element set is the time mini inertia measurement unit output information joint probability density function;

be illustrated in hypothesis H ₀under condition, the unknown signaling element set is

the time mini inertia measurement unit output information joint probability density function.

As a prioritization scheme of the embodiment of the present invention, in step 9, inequality is expressed as:

$T\left(z_n\right)=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}(\frac{1}{{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-g\frac{{\stackrel{-}{y}}_n^a}{\left|\right|{\stackrel{-}{y}}_n^a\left|\right|}\left|\right|}^2+\frac{1}{{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}\left|\right|}^2)<{\mathrm{\&gamma;}}^{\&prime;}$

If inequality is set up, individual soldier's autonomous navigation system zero-speed testing result is static-H ₁be true, otherwise the zero-speed testing result is motion-H ₀be true;

Wherein, ${\stackrel{-}{y}}_n^a=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}{y}_k^a,$ $T\left(z_n\right)=-\frac{2}{N}\ln L_G\left(z_n\right),$ γ '=-(2/N) ln (γ), ln (.) means to ask take the logarithm that e is the truth of a matter.

Below in conjunction with drawings and the specific embodiments, application principle of the present invention is further described.

As shown in Figure 1, the zero-speed detection method based on Nei Man-Pearson criterion of the embodiment of the present invention comprises the following steps:

S101: hand-held palm PC receives and stores in real time the measurement information of step mini inertia measurement unit output in individual soldier's autonomous navigation system;

S102: data transfer rate between mini inertia measurement unit and palm PC in mini inertia measurement unit length rest time and measuring process before using according to the sample frequency of each sensor in individual soldier's autonomous navigation system, system obtains the concrete value of window function N in this zero-speed testing process;

S103: utilize the mini inertia measurement unit output data and the window function N that collect, the zero-speed test problems is converted into to modeled pair of Hypothesis Testing Problem;

S104: the signal that utilizes mini inertia measurement unit sensor output signal characteristics, palm PC to receive and the character of error disturbing signal thereof, and formula, the mathematical model of acquisition mini inertia measurement unit inertial sensor output signal;

S105: mathematical model and the formula of the data that in individual soldier's autonomous navigation system motion process that utilization collects, mini inertia measurement unit is exported, definite mini inertia measurement unit inertial sensor output signal, the joint probability density function of acquisition sensor measurement amount;

S106: utilize the observation signal joint probability density function obtained, according to formula, obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test;

S107: utilize formula, the maximum likelihood that obtains unknown signaling element under two kinds of assumed conditions is estimated;

S108: use: the maximum likelihood of trying to achieve the unknown signaling element estimates, replaces unknown signaling element set in extensive probability likelihood ratio test and obtains not the extensive probability likelihood ratio containing unknown element;

S109: utilize the extensive probability likelihood ratio and the inequality that obtain, determine individual soldier's autonomous navigation system zero-speed testing result;

Concrete steps of the present invention are:

Step 1: hand-held palm PC receives and stores in real time the measurement information of step mini inertia measurement unit output in individual soldier's autonomous navigation system, and the total output information of the mini inertia measurement unit that any time k receives is $y_k={\left[\begin{array}{cc}{y}_k^a& y_k^{\mathrm{\&omega;}}\end{array}\right]}^T;$

Wherein, T means matrix transpose operation, $y_k^{\mathrm{\&omega;}}={\left[\begin{array}{ccc}{y}_k^{\mathrm{\&omega;x}}& y_k^{\mathrm{\&omega;y}}& y_k^{\mathrm{\&omega;z}}\end{array}\right]}^T$ Angular speed information for the output of micro mechanical system (Micro-Electro-MechanicalSystem) three-axis gyroscope; $y_k^a={\left[\begin{array}{ccc}{y}_k^{\mathrm{ax}}& y_k^{\mathrm{ay}}& y_k^{\mathrm{az}}\end{array}\right]}^T$ Ratio force information for the output of micro mechanical system three axis accelerometer;

Step 2: data transfer rate between mini inertia measurement unit and palm PC in mini inertia measurement unit length rest time and step 1 measuring process before using according to the sample frequency of each sensor in individual soldier's autonomous navigation system, system obtains the concrete value (N is integer) of window function N in this zero-speed testing process;

Step 3: the mini inertia measurement unit output data and the definite window function N of step 2 that utilize step 1 to collect, the zero-speed test problems is converted into to modeled pair of Hypothesis Testing Problem, testing result is hypothesis H ₀(mini inertia measurement unit motion), H ₁one of (mini inertia measurement unit is static), utilize Nei Man-Pearson came theorem to obtain the mathematical model that zero-speed detects judgement: known spurious alarm probability P _fAduring=α, if meet zero-speed judgement inequality, be

h ₁be true, individual soldier's autonomous navigation system zero-speed testing result is that the step mini inertia measurement unit is static; Function L (z _n) relevant with likelihood ratio, for z _neach value, it has meaned H ₁the hypothetical probabilities value is to H ₀the ratio of hypothetical probabilities value;

Wherein, spurious alarm probability P _fA={ H ₁∣ H ₀mean when mini inertia measurement unit is motion state it is judged as to the probability of stationary state; z _n={ y _kthe n+N-1k=n measured value that is mini inertia measurement unit in a period of time; The γ value is by formula

definition; p(z _n; H ₀), p (z _n; H ₁) mean respectively the probability density function of observation data under two kinds of hypothesis;

Step 4: the signal that utilizes mini inertia measurement unit sensor output signal characteristics, palm PC to receive and the character of error disturbing signal thereof, and formula:

y _k＝s _k(θ)+v _k

Obtain the mathematical model of mini inertia measurement unit inertial sensor output signal;

In formula, $s_k\left(\mathrm{\&theta;}\right)={\left[\begin{array}{cc}{s}_k^a\left(\mathrm{\&theta;}\right)& s_k^{\mathrm{\&omega;}}\left(\mathrm{\&theta;}\right)\end{array}\right]}^T,$ $v_k={\left[\begin{array}{cc}{v}_k^a& v_k^{\mathrm{\&omega;}}\end{array}\right]}^T,$ $s_k^a\left(\mathrm{\&theta;}\right)\&Element;{\mathrm{\&Omega;}}^3,$ $s_k^{\mathrm{\&omega;}}\left(\mathrm{\&theta;}\right)\&Element;{\mathrm{\&Omega;}}^3$ Mean respectively k mini inertia measurement unit measures constantly specific force and angular speed, θ means the set of unknown element in needs description signal;

with

the noise set that expression is relevant to accelerometer and gyroscope respectively, T means matrix transpose operation; Suppose that noise is independent, zero-mean Gaussian distribution, its covariance matrix is $C=E\left\{\begin{array}{cc}{v}_k& {v_k}^T\end{array}\right\}=\mathrm{diag}{\left[\begin{array}{cc}{{\mathrm{\&sigma;}}_a}^2{I}_3& {{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}}^2{I}_3\end{array}\right]}^T,$ I wherein ₃(0 ₃) mean 3 * 3 unit (zero) battle array; Diag[] the expression diagonal matrix; E{} means to ask expectation,

with

mean respectively micro-mechanical accelerometer and gyrostatic noise variance;

Step 5: the mathematical model and the formula that utilize mini inertia measurement unit inertial sensor output signal definite in the data, step 4 of the output of mini inertia measurement unit in the individual soldier's autonomous navigation system motion process collected in step 1:

$p(z_n,{\mathrm{\&theta;}}_i,H_i)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k;{\mathrm{\&theta;}}_i,H_i)$

$=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_i,H_i)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_i,H_i)$

Obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Wherein,

i=0,1; θ _ibe illustrated in hypothesis H _iunknown signaling element set under condition, Π means even to take advantage of; p(z _n; θ _i, H _i) be illustrated in and suppose H _iunder condition, the unknown signaling element set is θ _ithe time individual soldier autonomous navigation system output signal joint probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical accelerometer output signal probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical gyroscope output signal probability density function;

$p(y_k^a;{\mathrm{\&theta;}}_i,H_i)=\frac{1}{{\left(2\mathrm{\&pi;}{\mathrm{\&sigma;}}_a^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-s_k^a\left({\mathrm{\&theta;}}_i\right)\left|\right|}^2\}$

$p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_i,H_i)=\frac{1}{{\left(2\mathrm{\&pi;}{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}-s_k^{\mathrm{\&omega;}}\left({\mathrm{\&theta;}}_i\right)\left|\right|}^2\}$

Wherein, exp{.} means the exponential function of e; || .|| means to ask norm; σ 2a, σ 2 ω mean respectively accelerometer and gyrostatic noise variance;

Step 6: utilize the observation signal joint probability density function obtained in step 5

P(z _n; θ _i, H _i), according to formula: $L_g\left(z_n\right)=\frac{\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_1,H_1)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_1,H_1)}{\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_0,H_0)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_0,H_0)}$

Obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Wherein, at hypothesis H ₁conditioned signal s _k(θ) fully unknown, the set of unknown element is:

${\mathrm{\&theta;}}_0\&equiv;\{z_k{\}}_{k=n}^{n+N-1};$

At hypothesis H ₀it is unknown that condition only has than the direction of force vector, and the set of unknown element is:

θ ₁≡gu _n，

In formula, u _n∈ Ω _u, Ω _u={ u ∈ R ³: ‖ u ‖=1}; G is the local gravity vector;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical accelerometer output signal probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical accelerometer output signal probability density function;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical gyroscope output signal probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical gyroscope output signal probability density function;

Step 7: utilize formula:

${\widehat{\mathrm{\&theta;}}}^0=\arg \max \left(p(z_n;{\mathrm{\&theta;}}_0,H_0)\right)$

${\widehat{\mathrm{\&theta;}}}^1=\arg \max \left(p(z_n;{\mathrm{\&theta;}}_1,H_1)\right)$

The maximum likelihood that obtains unknown signaling element under two kinds of assumed conditions is estimated

In formula,

at hypothesis H ₁the maximum likelihood of condition unknown element is estimated;

at hypothesis H ₀under condition, the maximum likelihood of unknown element is estimated; p(z _n; θ ₁, H ₁) be illustrated in and suppose H ₁under condition, the unknown signaling element set is θ ₁the time mini inertia measurement unit output information joint probability density function; p(z _n; θ ₀, H ₀) be illustrated in and suppose H ₀under condition, the unknown signaling element set is θ ₀the time mini inertia measurement unit output information joint probability density function; The value of unknown element set when argmax () means to make in bracket to get maximal value;

$p(z_n;{\mathrm{\&theta;}}_1,H_1)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_1,H_1)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_1,H_1)$

$p(z_n,{\mathrm{\&theta;}}_0,H_0)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_0,H_0)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_0,H_0);$

Step 8: the maximum likelihood that uses step 7 to try to achieve the unknown signaling element is estimated

step of replacing six L _g(z _n) in the unknown signaling element set obtain not the extensive probability likelihood ratio L containing unknown element _g(z _n);

$L_G\left(z_n\right)=\frac{p(z_n;{\widehat{\mathrm{\&theta;}}}^1,H_1)}{p(z_n;{\widehat{\mathrm{\&theta;}}}^0,H_0)}$

In formula,

be illustrated in hypothesis H ₁under condition, the unknown signaling element set is

the time mini inertia measurement unit output information joint probability density function;

be illustrated in hypothesis H ₀under condition, the unknown signaling element set is

the time mini inertia measurement unit output information joint probability density function;

Step 9: utilize the L obtained in step 8 _g(z _n) and inequality:

$T\left(z_n\right)=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}(\frac{1}{{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-g\frac{{\stackrel{-}{y}}_n^a}{\left|\right|{\stackrel{-}{y}}_n^a\left|\right|}\left|\right|}^2+\frac{1}{{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}\left|\right|}^2)<{\mathrm{\&gamma;}}^{\&prime;}$

Determine individual soldier's autonomous navigation system zero-speed testing result: if inequality is set up, individual soldier's autonomous navigation system zero-speed testing result is static-H ₁be true, otherwise the zero-speed testing result is motion-H ₀be true;

Wherein, ${\stackrel{-}{y}}_n^a=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}{y}_k^a,$ $T\left(z_n\right)=-\frac{2}{N}\ln L_G\left(z_n\right),$ γ '=-(2/N) ln (γ), ln (.) means to ask take the logarithm that e is the truth of a matter;

In conjunction with following experiment, excellent beneficial effect of the present invention is described further:

Utilization is built true individual soldier's Navigation System Model from Kenzo axle inertial measurement cluster (integrated micro mechanical system three axle magnetometers, accelerometer, gyroscope), device parameter is as shown in table 1, detect individual soldier that verification experimental verification resolves based on two mini inertia measurement unit inertia reliability, practicality, the accuracy of zero-speed detection algorithm in true environment of navigating by rational zero-speed, the test scene is selected in Harbin Engineering University's military project playground of outdoor spaciousness

Table 1 is from grinding each sensor performance index of mini inertia measurement unit inertial measurement cluster

In experimentation, relative parameters setting is as follows:

Individual soldier freedom positioning system sample frequency: the 100Hz that navigates

Window function N:5

Micro mechanical system gyro standard deviation: σ _a=0.01m/s ²

Micro-mechanical accelerometer standard deviation: σ _g=0.1*pi/180rad/s

Parameter γ=0.3e ⁵

Before experiment starts, the tester carries out the system quiescence preheating of 15 minutes, the initialization of completion system at field experiment;

Test 1:

The tester carries out respectively the motions such as " walking (1.05m/s) ", " (2.14m/s) jogs ", " (3.50m/s) hurries up " on the track on rectangle playground, Real-time Collection the output data of preserving mini inertia measurement unit in experimentation, as a reference, the zero-speed interval that observation gyro output valve is drawn is considered as real zero-speed interval, is illustrated in figure 2 after the first walking of pedestrian the output valve of Y-axis gyro while jogging;

Experimental data is carried out to off-line analysis, for comparatively validate experimental result of the present invention, provide the testing result of the zero-speed detection method based on the N-P criterion of traditional variance threshold values zero-speed detection method and the present invention's proposition simultaneously, and contrasted with gyrostatic raw data, be illustrated in figure 3 after first walking the testing result of jogging frequently with detection method;

As can be seen from Figure 3, variance threshold values step static detection method zero-speed when jogging detects and lost efficacy, and the algorithm that the present invention mentions can detect the zero-speed interval exactly;

The output valve of Y-axis gyro when being illustrated in figure 4 the pedestrian and jogging and hurry up, the zero-speed testing result as shown in Figure 5, as can be seen from the figure when the pedestrian hurries up, variance threshold values step static detection method testing result lost efficacy, in above-mentioned experiment, the algorithm that the present invention mentions can detect the zero-speed interval exactly;

Test 2:

In order quantitatively to contrast the performance of the whole bag of tricks, we allow the pedestrian with different speed walkings or run 30 steps, wherein each step has a zero-speed interval, have 30 zero-speed intervals, the interval quantity of the zero-speed that the zero-speed detection method based on Nei Man-Pearson came (N-P) criterion that variance threshold values step zero-speed detection method and the present invention propose detects is as shown in table 2, when walking, two kinds of algorithms can detect the zero-speed interval accurately; When running, variance threshold values zero-speed detection method lost efficacy, and the quantity that all can detect exactly the zero-speed interval is walked or run to algorithm of the present invention no matter,

The interval number of the zero-speed that table 2 distinct methods detects is (the interval number of actual zero-speed is 30) relatively

Table 3 has provided the error-detecting quantity in various algorithm zero-speeds interval, be about to the quantity that the interval wrong detection of non-zero-speed is the zero-speed interval, as can be seen from Table 3, variance threshold values zero-speed detection method detects the zero-speed interval sometimes mistakenly, and no matter the N-P zero-speed detection algorithm that the present invention mentions is that walking state or running state all can accurately and reliably detect the zero-speed interval, the zero-speed of this algorithm for normal walking with while running detects all effective, and reliability is high, good stability;

The number in table 3 distinct methods error-detecting zero-speed interval relatively

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.

Claims (7)

1. the zero-speed detection method based on Nei Man-Pearson criterion, is characterized in that, should the zero-speed detection method based on Nei Man-Pearson criterion comprise the following steps:

Step 1, hand-held palm PC receives and stores in real time the measurement information of step mini inertia measurement unit output in individual soldier's autonomous navigation system;

Step 2: data transfer rate between mini inertia measurement unit and palm PC in mini inertia measurement unit length rest time and step 1 measuring process before using according to the sample frequency of each sensor in individual soldier's autonomous navigation system, system, obtain the concrete value of window function N in this zero-speed testing process, N is integer;

Step 3: the mini inertia measurement unit output data and the definite window function N of step 2 that utilize step 1 to collect, the zero-speed test problems is converted into to modeled pair of Hypothesis Testing Problem, suppose that testing result is mini inertia measurement unit motion H ₀, the static H of mini inertia measurement unit ₁one of, utilize Nei Man-Pearson came theorem to obtain the mathematical model that zero-speed detects judgement: known spurious alarm probability P _fAduring=α, if meet zero-speed judgement inequality, be

h ₁be true, individual soldier's autonomous navigation system zero-speed testing result is that mini inertia measurement unit is static; Function L (z _n) relevant with likelihood ratio, for z _neach value, meaned H ₁the hypothetical probabilities value is to H ₀the ratio of hypothetical probabilities value;

Wherein, spurious alarm probability P _fA={ H ₁∣ H ₀while meaning that mini inertia measurement unit is motion state, the probability that the zero-speed testing result is stationary state; z _n={ y _kthe n+N-1k=n measured value that is mini inertia measurement unit in a period of time; The γ value is by formula

definition; p(z _n; H ₀), p (z _n; H ₁) mean respectively the probability density function of observation data under two kinds of hypothesis;

Step 4: the signal that utilizes mini inertia measurement unit sensor output signal characteristics, palm PC to receive and the character of error disturbing signal, and formula, the mathematical model of acquisition mini inertia measurement unit inertial sensor output signal;

Step 5: utilize mathematical model and the formula of mini inertia measurement unit inertial sensor output signal definite in the data, step 4 of the output of mini inertia measurement unit in the individual soldier's autonomous navigation system motion process collected in step 1, obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Step 6: utilize the observation signal joint probability density function p (z obtained in step 5 _n; θ _i, H _i), according to formula, obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Step 7: utilize formula:

${\widehat{\mathrm{\&theta;}}}^0=\arg \max \left(p(z_n;{\mathrm{\&theta;}}_0,H_0)\right)$

${\widehat{\mathrm{\&theta;}}}^1=\arg \max \left(p(z_n;{\mathrm{\&theta;}}_1,H_1)\right)$

The maximum likelihood that obtains unknown signaling element under two kinds of assumed conditions is estimated

In formula,

at hypothesis H ₁the maximum likelihood of condition unknown element is estimated;

at hypothesis H ₀under condition, the maximum likelihood of unknown element is estimated; p(z _n; θ ₁, H ₁) be illustrated in and suppose H ₁under condition, the unknown signaling element set is θ ₁the time mini inertia measurement unit output information joint probability density function; p(z _n; θ ₀, H ₀) be illustrated in and suppose H ₀under condition, the unknown signaling element set is θ ₀the time mini inertia measurement unit output information joint probability density function; The value of unknown element set when argmax () means to make in bracket to get maximal value;

$p(z_n;{\mathrm{\&theta;}}_1,H_1)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_1,H_1)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_1,H_1)$

$p(z_n,{\mathrm{\&theta;}}_0,H_0)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_0,H_0)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_0,H_0);$

Step 8: the maximum likelihood that uses step 7 to try to achieve the unknown signaling element is estimated

step of replacing six L _g(z _n) in the unknown signaling element set obtain not the extensive probability likelihood ratio L containing unknown element _g(z _n);

Step 9: utilize the L obtained in step 8 _g(z _n) and inequality, determine individual soldier's autonomous navigation system zero-speed testing result.

2. the zero-speed detection method based on Nei Man-Pearson criterion as claimed in claim 1, is characterized in that, in step 1, the total output information of the mini inertia measurement unit that any time k receives is $y_k={\left[\begin{array}{cc}{y}_k^a& y_k^{\mathrm{\&omega;}}\end{array}\right]}^T;$

Wherein, T means matrix transpose operation, $y_k^{\mathrm{\&omega;}}={\left[\begin{array}{ccc}{y}_k^{\mathrm{\&omega;x}}& y_k^{\mathrm{\&omega;y}}& y_k^{\mathrm{\&omega;z}}\end{array}\right]}^T$ Angular speed information for the output of micro mechanical system three-axis gyroscope; $y_k^a={\left[\begin{array}{ccc}{y}_k^{\mathrm{ax}}& y_k^{\mathrm{ay}}& y_k^{\mathrm{az}}\end{array}\right]}^T$ Ratio force information for the output of micro mechanical system three axis accelerometer.

3. the zero-speed detection method based on Nei Man-Pearson criterion as claimed in claim 1, is characterized in that, in step 4, passes through formula:

y _k＝s _k(θ)+v _k

Obtain the mathematical model of mini inertia measurement unit inertial sensor output signal;

In formula, $s_k\left(\mathrm{\&theta;}\right)={\left[\begin{array}{cc}{s}_k^a\left(\mathrm{\&theta;}\right)& s_k^{\mathrm{\&omega;}}\left(\mathrm{\&theta;}\right)\end{array}\right]}^T,$ $v_k={\left[\begin{array}{cc}{v}_k^a& v_k^{\mathrm{\&omega;}}\end{array}\right]}^T,$ $s_k^a\left(\mathrm{\&theta;}\right)\&Element;{\mathrm{\&Omega;}}^3$ 、 $s_k^{\mathrm{\&omega;}}\left(\mathrm{\&theta;}\right)\&Element;{\mathrm{\&Omega;}}^3$ Mean respectively k mini inertia measurement unit measures constantly specific force and angular speed, θ means the set of unknown element in needs description signal;

with

the noise set that expression is relevant to accelerometer and gyroscope respectively, T means matrix transpose operation; Suppose that noise is independent, zero-mean Gaussian distribution, covariance matrix is $C=E\left\{\begin{array}{cc}{v}_k& {v_k}^T\end{array}\right\}=\mathrm{diag}{\left[\begin{array}{cc}{{\mathrm{\&sigma;}}_a}^2{I}_3& {{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}}^2{I}_3\end{array}\right]}^T,$ I wherein ₃(0 ₃) mean 3 * 3 unit (zero) battle array; Diag[] the expression diagonal matrix; E{} means to ask expectation,

with

mean respectively micro-mechanical accelerometer and gyrostatic noise variance.

4. the zero-speed detection method based on Nei Man-Pearson criterion as claimed in claim 1, is characterized in that, in step 5, passes through formula:

$p(z_n,{\mathrm{\&theta;}}_i,H_i)=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k;{\mathrm{\&theta;}}_i,H_i)$

$=\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_i,H_i)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_i,H_i)$

Obtain the joint probability density function p (z of sensor measurement amount _n; θ _i, H _i);

Wherein,

i=0,1; θ _ibe illustrated in hypothesis H _iunknown signaling element set under condition, Π means even to take advantage of; p(z _n; θ _i, H _i) be illustrated in and suppose H _iunder condition, the unknown signaling element set is θ _ithe time individual soldier autonomous navigation system output signal joint probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical accelerometer output signal probability density function;

mean k constantly, at hypothesis H _iunder condition, the unknown signaling element set is θ _ithe time micro-mechanical gyroscope output signal probability density function;

$p(y_k^a;{\mathrm{\&theta;}}_i,H_i)=\frac{1}{{\left(2\mathrm{\&pi;}{\mathrm{\&sigma;}}_a^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-s_k^a\left({\mathrm{\&theta;}}_i\right)\left|\right|}^2\}$

$p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_i,H_i)=\frac{1}{{\left(2\mathrm{\&pi;}{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2\right)}^{3/2}}\exp \{-\frac{1}{2{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}-s_k^{\mathrm{\&omega;}}\left({\mathrm{\&theta;}}_i\right)\left|\right|}^2\}$

Wherein, exp{.} means the exponential function of e; || || mean to ask norm; σ 2a, σ 2 ω mean respectively accelerometer and gyrostatic noise variance.

5. the zero-speed detection method based on Nei Man-Pearson criterion as claimed in claim 1, is characterized in that,

In step 6, according to formula: $L_g\left(z_n\right)=\frac{\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_1,H_1)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_1,H_1)}{\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Pi;}}p(y_k^a;{\mathrm{\&theta;}}_0,H_0)p(y_k^{\mathrm{\&omega;}};{\mathrm{\&theta;}}_0,H_0)}$

Obtain individual soldier's autonomous navigation system zero-speed that contains unknown element and detect extensive probability likelihood ratio test L _g(z _n);

Wherein, at hypothesis H ₁conditioned signal s _k(θ) fully unknown, the set of unknown element is:

${\mathrm{\&theta;}}_0\&equiv;\{z_k{\}}_{k=n}^{n+N-1};$

At hypothesis H ₀it is unknown that condition only has than the direction of force vector, and the set of unknown element is:

θ ₁≡gu _n，

In formula, u _n∈ Ω _u, Ω _u={ u ∈ R ³: ‖ u ‖=1}; G is the local gravity vector;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical accelerometer output signal probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical accelerometer output signal probability density function;

be engraved in hypothesis H while meaning k ₁under condition, the unknown signaling element set is θ ₁the time micro-mechanical gyroscope output signal probability density function;

be engraved in hypothesis H while meaning k ₀under condition, the unknown signaling element set is θ ₀the time micro-mechanical gyroscope output signal probability density function.

6. the zero-speed detection method based on Nei Man-Pearson criterion as claimed in claim 1, is characterized in that, in step 8, and extensive probability likelihood ratio L _g(z _n) be expressed as;

$L_G\left(z_n\right)=\frac{p(z_n;{\widehat{\mathrm{\&theta;}}}^1,H_1)}{p(z_n;{\widehat{\mathrm{\&theta;}}}^0,H_0)}$

In formula,

be illustrated in hypothesis H ₁under condition, the unknown signaling element set is

the time mini inertia measurement unit output information joint probability density function;

be illustrated in hypothesis H ₀under condition, the unknown signaling element set is the time mini inertia measurement unit output information joint probability density function.

7. the zero-speed detection method based on Nei Man-Pearson criterion as claimed in claim 1, is characterized in that, in step 9, inequality is expressed as:

$T\left(z_n\right)=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}(\frac{1}{{\mathrm{\&sigma;}}_a^2}{\left|\right|y_k^a-g\frac{{\stackrel{-}{y}}_n^a}{\left|\right|{\stackrel{-}{y}}_n^a\left|\right|}\left|\right|}^2+\frac{1}{{\mathrm{\&sigma;}}_{\mathrm{\&omega;}}^2}{\left|\right|y_k^{\mathrm{\&omega;}}\left|\right|}^2)<{\mathrm{\&gamma;}}^{\&prime;}$

If inequality is set up, individual soldier's autonomous navigation system zero-speed testing result is static-H ₁be true, otherwise the zero-speed testing result is motion-H ₀be true;

Wherein, ${\stackrel{-}{y}}_n^a=\frac{1}{N}\underset{k\&Element;{\mathrm{\&Omega;}}_n}{\mathrm{\&Sigma;}}{y}_k^a,$ $T\left(z_n\right)=-\frac{2}{N}\ln L_G\left(z_n\right),$ γ '=-(2/N) ln (γ), ln (.) means to ask take the logarithm that e is the truth of a matter.

Images