CN102944216A - Three-redundant ship dynamic positioning heading measurement method based on improved voting algorithm - Google Patents

Abstract

The invention relates to a three-redundant ship dynamic positioning heading measurement method based on an improved voting algorithm. The method comprises the following steps: 1, measuring a heading value: 2, adopting a variance self-learning weighting method to obtain respective weight values; 3, carrying out weighting on the measured mean, respectively subtracting a weighted mean from an input value of a sensor, calculating an absolute value, and comparing the absolute value with a threshold t1, wherein the absolute value less than the threshold t1 is considered that no failure exists so as to directly perform data fusion, and it is considered that the failure exists if the absolute value greater than t1 exists, such that failure redetermination is required to be performed; 4, obtaining difference between the current input value and the previous output value, and calculating an absolute value, wherein a normal state is determined when the absolute value is less than t2 and a heading change rate at the current time is similar to heading change rates of the current rudder angle and the ship speed, and otherwise a failure is determined; and 5, carrying out weighting on the remaining data reassignment weight values after excluding the failure data. With the present invention, a lot of failure determinations are achieved, such that the output result has high precision.

Description

Based on the triple redundance dynamic positioning of vessels bow that improves voting algorithm to measuring method

Technical field

The present invention relates to the dynamic positioning of vessels device, relate in particular to a kind of triple redundance dynamic positioning of vessels bow based on improving voting algorithm to measuring method.

Background technology

Bow to measuring system be used for to measure in the ship's navigation ship bow to the system of geographic north to angle.Existing bow generally includes to system: (1) gyro compass, be used for to measure bow to; (2) magnetic compass: as for subsequent use measurement mechanism of bow to measuring system; (3) central computer, the bow that gyro compass is recorded carries out fault judgement and Data Fusion to data.Generally, the bow of boats and ships only is equipped with a cover gyro compass usually to measuring system, is used to bow to provide data to measurement.Because bow is to having key effect for Ship Dynamic Positioning Systems Based, the mistake bow is to a series of serious consequences such as will cause that vessel position is lost.Therefore, International Maritime Organization (IMO) and the classification societies all stipulate to install simultaneously three bows to sensor for 3 grades of dynamic positioning systems.The redundant bow that consists of to sensor for three bows is to measuring system, in the time of should guaranteeing that wherein single bow is to sensor fault, system should to fault sensor judge and provide remove fault effects bow to metrical information.Therefore, the reliability of measuring to measuring system in order to improve bow is equipped with above-mentioned existing bow single cover of gyro compass in the measuring system, rises to three cover gyro compasses and is equipped with, namely have three cover gyro compasses to be used for measurement data, be commonly referred to: the bow of triple redundance type dynamic positioning system is to measuring system.The normal method that adopts is to adopt voting algorithm that bow is processed to metrical information at the triple redundance bow in measuring system.

As shown in Figure 1, it is the structure that the bow of a typical triple redundance dynamic positioning system adopts to measuring system, it is made of three gyro compasses 1, three gyro compasses 1 are measured respectively bow to data, then, gyro compass 1 passes through serial communication interface, the ship bow that measures is sent to central computer 2 to information, central computer 2 is by data acquisition and handling procedure control data acquisition port, receive in real time the measurement data from gyro compass 1, and with measurement data process majority votinl and the data fusion received, in the fusion process, the bow that calculates by the voting module 3 in the central computer 2 and weighting block 4 is worth to fusion, and each sensor is endowed identical weights usually, exports at last bow to value.

After adopting this redundancy, bow increases to the more single compass measurement of the reliability of measuring system, but above-mentioned redundancy structure still has the following disadvantages:

(1) because in the system, the weights of each Data Fusion of Sensor distribute not have abundant performance difference in each sensor actual measurement, the fusion accuracy space that also improves a lot.

(2) the majority votinl mode that adopts can't accurately be implemented in break down the fault voting in the situation of two gyro compasses.

Summary of the invention

Fundamental purpose of the present invention is to overcome the above-mentioned shortcoming that prior art exists, and provides a kind of triple redundance dynamic positioning of vessels bow based on improving voting algorithm to measuring method, and it can realize most faults judgements, so that Output rusults has higher precision.

The objective of the invention is to be realized by following technical scheme:

A kind of triple redundance dynamic positioning of vessels bow based on improving voting algorithm is characterized in that to measuring method: adopt following steps:

The first step: measure bow to value to measuring unit by bow;

Second step: the bow of bow being measured gained to measuring unit by the weights allocation units is to value, and the method by variance self study weighting obtains weights separately again;

The 3rd step: by fault just judging unit bow be weighted to the measured value of measuring unit try to achieve the measurement average

, and the input value of three sensors deducted respectively weighted mean value

, it is asked to thoroughly deserve E1, E2, E3 again, then, E1, E2, E3 and threshold value t1 are made comparisons, if all less than threshold value t1, then think directly to enter non-fault the data fusion unit and carry out data fusion; If any one is arranged greater than t1 among E1, E2, the E3, then think and break down, need to carry out fault and judge again;

The 4th step: by fault again judging unit input value that sensor is current and the output valve of previous moment system are done difference respectively, take absolute value again, if absolute value is less than predetermined threshold value t2, and the bow surveyed of nearest a period of time of sensor will be close to rate of change with the bow that current rudder angle and the speed of a ship or plane determine to rate of change, thinks that then this sensor is normal; Otherwise, be judged as fault;

The 5th step: after the data eliminating of data fusion unit with fault, remaining data are redistributed weights be weighted.

Described bow is made of three gyro compasses to measuring unit.

Described weights allocation units adopt a kind of Multisensor Data Fusion Algorithm based on variance self study weighting, and this algorithm is treated old measurement data and given different weight coefficients with new measurement data, makes the process that variance progressively is tending towards optimum of measuring.

Described fault is just decided by vote the unit and is by gyro compass input data, weighted mean module, makes the differential mode piece and threshold value judges that four parts are composed in series successively.

It is to obtain the gyro compass measurement data in the unit by just deciding by vote in fault that described fault is decided by vote the unit again, try to achieve the measurement variation rate of each gyro compass, and bow in the mathematical model of itself and ship motion compared to rate of change, try to achieve difference, judge that by threshold value II realizes that fault judges again, the information of comparison be that the bow of the current value rate of change of each gyro compass and the decision of current ship motion model is to rate of change.

Weights in the described second step are

$\mathrm{Wi}\left(k\right)=\left(\frac{1}{\stackrel{\‾}{\mathrm{Ri}}\left(k\right)}\right)\frac{1}{{\mathrm{\Σ}}_{i=1}^n\frac{1}{\stackrel{\‾}{\mathrm{Ri}}\left(k\right)}}$

Wherein, The estimated value of i sensor-measurement variance when being the k time sampling.

Described fault also is provided with fault auxiliary judgment unit in the judging unit again, and fault auxiliary judgment unit is as the mathematical model of the kinematic parameter speed of a ship or plane and rudder angle and the ship motion of ship motion process, is used for auxiliaryly carrying out fault and deciding by vote.

Beneficial effect of the present invention: the present invention is owing to adopt technique scheme, and it can realize most faults judgements, so that Output rusults has higher precision.

Description of drawings:

Fig. 1 is for having the triple redundance bow now to the measurement mechanism structural drawing.

Fig. 2 fault of the present invention is just decided by vote the cellular construction synoptic diagram.

Fig. 3 is that fault of the present invention is decided by vote the cellular construction synoptic diagram again.

Fig. 4 is schematic flow sheet of the present invention.

Fig. 5 is measurement of the present invention and fused data figure.

Major label description in figure:

1. gyro compass, 2. central computer, 3. voting module, 4. weighting block, 5. gyro compass is inputted data, 6. weighted mean module, 7. make the differential mode piece, 8. threshold value is judged, 9. gyro compass measurement data, 10. gyro compass measurement variation rate, 11. bow is to rate of change, 12. difference, 13. threshold value is judged, 14. bow is to measuring unit, 15. weights allocation units, 16. fault is just decided by vote the unit, 17. auxiliary judgment unit, 18. fault is decided by vote the unit again, 19. data fusion unit.

Embodiment

Such as Fig. 2-shown in Figure 4, the present invention includes: bow is to measuring unit 14, weights allocation units 15, fault is just decided by vote unit 16, auxiliary judgment unit 17, and fault is decided by vote unit 18 and data fusion unit 19 again, wherein, bow is to be made of three gyro compasses 1 to measuring unit 14; Weights allocation units 15 adopt a kind of Multisensor Data Fusion Algorithm based on variance self study weighting, it is treated old measurement data and gives different weight coefficients with new measurement data, when considering that most recent data is brought into play Main Function in estimation, take full advantage of empirical data in the past, thereby, make the process that variance progressively is tending towards optimum of measuring; Fault is just decided by vote unit 16 and is by gyro compass input data 5, weighted mean module 6, makes differential mode piece 7 and threshold value judges that 8 four parts are composed in series (as shown in Figure 2) successively.

It is to obtain gyro compass measurement data 9 in the unit 16 by just deciding by vote in fault that fault is decided by vote unit 18 again, try to achieve the measurement variation rate 10 of each gyro compass 1, and bow in the mathematical model of itself and ship motion compared to rate of change 11, try to achieve difference 12, judge that by second threshold value 13 realize that faults judge again, the information of comparison be that the bow of the current value rate of change of each gyro compass 1 and the decision of current ship motion model is to rate of change 11.(as shown in Figure 3).

Specific implementation of the present invention comprises following step:

The first step: bow is to measurement: measure bow to value by bow to three gyro compasses 1 in the measuring unit 14;

Second step: weights distribute: the bow of bow each gyro compass 1 in the measuring unit 14 being measured gained by weights allocation units 15 is to value, and the method by variance self study weighting obtains weights separately again

$\mathrm{Wi}\left(k\right)=\left(\frac{1}{\stackrel{\‾}{\mathrm{Ri}}\left(k\right)}\right)\frac{1}{{\mathrm{\Σ}}_{i=1}^n\frac{1}{\stackrel{\‾}{\mathrm{Ri}}\left(k\right)}}$

Wherein,

The estimated value of i sensor-measurement variance when being the k time sampling.

The 3rd step: fault is just judged: fault just judging unit 16 is weighted bow measured value of each gyro compass 1 in the measuring unit 14 and tries to achieve the measurement average , and the input value of three sensors deducted respectively weighted mean value

, it is asked to thoroughly deserve E1, E2, E3 again, then, E1, E2, E3 and threshold value t1 are made comparisons, if all less than threshold value t1, then think directly to enter non-fault the data fusion unit and carry out data fusion; If any one is arranged greater than t1 among E1, E2, the E3, then think and break down, then need to carry out fault and judge again;

The 4th step: fault is judged again: fault again judging unit 18 input value that sensor is current and the output valve of previous moment system are done difference respectively, take absolute value again, if absolute value is less than predetermined threshold value t2, and the bow surveyed of nearest a period of time of sensor will be close to rate of change with the bow that current rudder angle and the speed of a ship or plane determine to rate of change, thinks that then this sensor is normal; Otherwise, be judged as fault; In addition, also be provided with fault auxiliary judgment unit 17 in the fault auxiliary judgment unit 18 as the mathematical model of the kinematic parameter speed of a ship or plane and rudder angle and the ship motion of ship motion process, be used for auxiliaryly carrying out fault and deciding by vote again;

The 5th step: data fusion: data fusion unit 19 is redistributed weights to remaining data and is weighted after the data of fault are got rid of.For example, when sensor 1 was judged to be fault, this moment, the weights of sensor 2 became W2 '=W2/ (W2+W3), and the weights of sensor 3 then are W3 '=W3/ (W2+W3), and the fusion output valve of this moment is (as shown in Figure 4).

Embodiment: as shown in Figure 5, under laboratory environment, the gyration of simulation boats and ships is carried out the emulation experiment of two gyro compass faults to bow to system, and adopts above-mentioned steps to carry out data and process.For merging the output data, Fig. 5 a, Fig. 5 b, Fig. 5 c are the measurement data of gyro compass 1 among its result such as Fig. 5 d.

The present invention can guarantee the fault Accuracy of Judgement owing to adopt to improve weighting algorithm and multistage fault judgment mechanism, and have that measuring error is little, regulating power strong, the advantages such as real-time and ease for use.

The above, it only is preferred embodiment of the present invention, be not that the present invention is done any pro forma restriction, every foundation technical spirit of the present invention all still belongs in the scope of technical solution of the present invention any simple modification, equivalent variations and modification that above embodiment does.

Claims (7)

1. One kind based on the triple redundance dynamic positioning of vessels bow that improves voting algorithm to measuring method, it is characterized in that: adopt following steps:

   The first step: measure bow to value to measuring unit by bow;

   Second step: the bow of bow being measured gained to measuring unit by the weights allocation units is to value, and the method by variance self study weighting obtains weights separately again;

   The 3rd step: by fault just judging unit bow be weighted to the measured value of measuring unit try to achieve the measurement average

   

   , and the input value of three sensors deducted respectively weighted mean value

   

   , it is asked to thoroughly deserve E1, E2, E3 again, then, E1, E2, E3 and threshold value t1 are made comparisons, if all less than threshold value t1, then think directly to enter non-fault the data fusion unit and carry out data fusion; If any one is arranged greater than t1 among E1, E2, the E3, then think and break down, need to carry out fault and judge again;

   The 4th step: by fault again judging unit input value that sensor is current and the output valve of previous moment system are done difference respectively, take absolute value again, if absolute value is less than predetermined threshold value t2, and the bow surveyed of nearest a period of time of sensor will be close to rate of change with the bow that current rudder angle and the speed of a ship or plane determine to rate of change, thinks that then this sensor is normal; Otherwise, be judged as fault;

   The 5th step: after the data eliminating of data fusion unit with fault, remaining data are redistributed weights be weighted.
2. 2. the triple redundance dynamic positioning of vessels bow based on improving voting algorithm according to claim 1 is to measuring method, and it is characterized in that: described bow is made of three gyro compasses to measuring unit.
3. 3. the triple redundance dynamic positioning of vessels bow based on improving voting algorithm according to claim 1 is to measuring method, it is characterized in that: described weights allocation units adopt a kind of Multisensor Data Fusion Algorithm based on variance self study weighting, this algorithm is treated old measurement data and is given different weight coefficients with new measurement data, makes the process that variance progressively is tending towards optimum of measuring.
4. 4. the triple redundance dynamic positioning of vessels bow based on improving voting algorithm according to claim 1 is characterized in that to measuring method: described fault is just decided by vote the unit and is by gyro compass input data, weighted mean module, makes the differential mode piece and threshold value judges that four parts are composed in series successively.
5. 5. the triple redundance dynamic positioning of vessels bow based on improving voting algorithm according to claim 1 is to measuring method, it is characterized in that: it is to obtain the gyro compass measurement data in the unit by just deciding by vote in fault that described fault is decided by vote the unit again, try to achieve the measurement variation rate of each gyro compass, and bow in the mathematical model of itself and ship motion compared to rate of change, try to achieve difference, judge that by threshold value II realizes that fault judges again, the information of comparison be that the bow of the current value rate of change of each gyro compass and the decision of current ship motion model is to rate of change.
6. 6. the triple redundance dynamic positioning of vessels bow based on improving voting algorithm according to claim 1 is to measuring method, and it is characterized in that: the weights in the described second step are

   $\mathrm{Wi}\left(k\right)=\left(\frac{1}{\stackrel{\‾}{\mathrm{Ri}}\left(k\right)}\right)\frac{1}{{\mathrm{\Σ}}_{i=1}^n\frac{1}{\stackrel{\‾}{\mathrm{Ri}}\left(k\right)}}$

   Wherein,

   

   The estimated value of i sensor-measurement variance during i.e. the k time sampling.
7. 7. the triple redundance dynamic positioning of vessels bow based on improving voting algorithm according to claim 1 is to measuring method, it is characterized in that: described fault also is provided with fault auxiliary judgment unit in the judging unit again, fault auxiliary judgment unit is as the mathematical model of the kinematic parameter speed of a ship or plane and rudder angle and the ship motion of ship motion process, is used for auxiliaryly carrying out fault and deciding by vote.

Images