CN102012242B - Fault-tolerant measurement method based on double queues - Google Patents
Fault-tolerant measurement method based on double queues Download PDFInfo
- Publication number
- CN102012242B CN102012242B CN201010521783XA CN201010521783A CN102012242B CN 102012242 B CN102012242 B CN 102012242B CN 201010521783X A CN201010521783X A CN 201010521783XA CN 201010521783 A CN201010521783 A CN 201010521783A CN 102012242 B CN102012242 B CN 102012242B
- Authority
- CN
- China
- Prior art keywords
- data
- measurement
- queue
- validity
- candidate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Testing Or Calibration Of Command Recording Devices (AREA)
- Complex Calculations (AREA)
Abstract
The invention belongs to the technical field of automatic control, in particular to a fault-tolerant measurement method based on double queues for external physical quantity. The method is as follows: double queues are adopted for storing measuring data, and measurement is carried out according to methods of effective judgment and compatible judgment. The measurement of the physical quantity is carried out by utilizing a high-accuracy rapid multiple- sampling averaging method so as to eliminate random errors, and improve the measurement accuracy. A measured result is subject to validity judgment before entering the queues, and whether the result is valid is judged according to the set dynamic random error range so as to eliminate the interference of one-time random events. When the interferences of twice random events occur continuously and are recorded by sampling queues, the measuring result enters a candidate queue, and finally a correct measuring result is given. By utilizing the measurement method, the random errors with normal distribution characteristics can be eliminated, the primary random errors can be eliminated, and the physical quantity under the condition of tolerating continuous twice random errors is correctly measured. The method can be widely applicable to measurement of the physical quantity with the dynamic random error characteristic.
Description
Technical field
The invention belongs to the automatic control technology field, be specifically related to a kind of fault-tolerant measuring method based on deque to outside physical quantity.
Background technology
Correct measurement is the basis of control system normal and stable work to outside physical quantity.In real system, the measurement of any physical quantity tends to be attended by the interference of dynamic random sum of errors random occurrence.Can not correct handling if these sum of errors disturb, the measurement result that will lead to errors and the control system run-time error occurs.For improving the accuracy of physical quantity, mainly adopted three kinds of improved measuring methods in the existing control system:
First kind is the precision that improves metering circuit, such as adopting more high-precision change-over circuit etc.This method can effectively improve because the precise decreasing that causes of stochastic error, but because metering circuit is not judged the validity of measurement result, so stochastic error and random error still be present in the measurement result, causes result's validity to reduce greatly.
Second kind then is repeatedly to measure and get average fast.This method can be eliminated the interference that stochastic error with normal distribution characteristic is brought, and removes mean value through repeatedly measuring, and makes a plurality of dynamic random errors eliminate the effects of the act each other, obtains accurate measurement result.But this method can not be eliminated the influence of random external incident to measurement result; Excessive or the too small situation of measurement result that causes such as the random external incident; This method is only through average; Stochastic events is reduced by half, when random occurrence occurs, cause the mistake of measurement result or error excessive easily.
The third then is an interval sampling, and sampled data is removed maximin, gets average then.This method can reduce because random occurrence causes the probability of measurement result mistake, causes the excessive or too small situation of measurement result can obtain correct measurement result to a random occurrence.But in the measuring process repeatedly at interval disposable random occurrence can't handle, the double situation that random occurrence occurs then can't be got rid of influence more, obtains correct measurement result.
More than three kinds of improved measuring methods all can not well get rid of stochastic error and random occurrence interference to measuring process and measurement result.And when in the control system of reality, carrying out physical quantity, the interference of stochastic error and random occurrence exists all the time.The size of stochastic error is because different physical quantity characteristic and metering circuit characteristic and difference, and the probability and the frequency of random occurrence appearance also have nothing in common with each other.For elimination has the stochastic error of normal distribution characteristic, and under the situation that the random occurrence interference occurs, obtain correct measurement result, need a kind of fault-tolerant measuring method of science.
Summary of the invention
The objective of the invention is to provides a kind of measuring method based on deque to existing above-mentioned defective in the present technology.
The inventive method comprises the step of following order:
(1) at first, to different physical quantitys, set four parameters:
(a) the gap periods T that measures, the time of promptly repeatedly waiting between the measure physical quantities;
(b) the scope W of stochastic error promptly is used for confirming the error range of measurement data validity;
(c) the definite times N of measurement result is promptly confirmed the required effective measurement number of times of measurement result;
(d) the maximum number of times M that measures before promptly measurement result is confirmed, allows the number of times of maximum measure physical quantities;
(2) after the measurement beginning, physical quantity is accurately measured, each measurement result is handled as follows:
One of which, measurement result at first carry out sampling queue to be handled:
(a) preceding twice measurement result directly got into sampling queue;
(b) follow-up measurement result is at first according to the error margin W that sets, and the measurement result in the sampling queue, judges the validity of this measurement data;
(c) if be judged as valid data, then directly get into sampling queue, the valid data number is added 1; If the valid data number reaches the parameter N of setting, then finish to measure, the valid data in the sampling queue are got average return;
(d) if be not sure of its validity, then judge the compatibility of data in these data and the sampling queue, as if with sampling queue in data compatibility, then upgrade the measurement data in the sampling queue;
Its two, for can't judgment data validity, with the incompatible measurement data of existing sampling queue data, carry out candidate's measurement processing again:
(a) preceding twice candidate's measurement result directly got into candidate queue;
(b) follow-up candidate's measurement result is at first according to the error margin W that sets, and data are carried out the validity judgement in the candidate queue;
(c) if be judged as valid data, then get into candidate queue, candidate's valid data number is added 1; If candidate's valid data number reaches the parameter N of setting, then finish to measure, the valid data in the candidate queue are got average return;
(d) if validity that still can not judgment data, then judge the compatibility of data in these data and the candidate queue, as if with candidate queue in data compatibility, then upgrade the measurement data in the candidate queue;
Its three, if be not sure of the validity of measurement data, measurement data is compatible with said sampling queue and candidate queue again, judges then whether the measurement number of times reaches setup parameter M:
(a) if reach setup parameter M, then finish to measure, get average according to the more queuing data of valid data in sampling queue and the candidate queue, confirm to measure the result;
(b), measure once more after then waiting for interval time of setup parameter T if do not reach setup parameter M.
This method adopts deque to deposit measurement data, measures according to the method for validity judgement and compatibility determination.The method that the measurement of physical quantity adopts the quick repeatedly sampling of high precision to get average is eliminated stochastic error, improves the measurement result precision.Measurement result is carried out the validity judgement before getting into formation, and according to the dynamic random error range of setting, whether judged result is effective, is used to eliminate a random occurrence and disturbs.When double random occurrence interference occurring and having been write down by sampling queue, measurement result gets into the candidate formation, and finally provides correct measurement result.This measuring method can be eliminated the stochastic error of normal distribution characteristic, gets rid of one time random error, and is allowing correct measurement physical quantity under the situation of double random error.The inventive method can generally be applicable to the measurement of the physical quantity with dynamic random error characteristics.
Description of drawings
Fig. 1 is the process flow diagram of the inventive method.
Embodiment
Below in conjunction with accompanying drawing and embodiment the present invention is described in further detail.
This measuring method through to physical quantity at interval property repeatedly measure, and the method for getting average obtains physical quantity result accurately, measuring process is as shown in Figure 1.To different physical quantities, need to set four parameters: the gap periods T of measurement, the scope W of stochastic error, and the times N that measurement result is confirmed is measured number of times M with maximum.The time of gap periods for repeatedly waiting between the measure physical quantities.The scope W of stochastic error is used for confirming the validity of measurement result.The times N that measurement result is confirmed is meant confirms the required effective measurement number of times of measurement result.Before maximum measurement number of times is meant that measurement result is confirmed, allow the number of times of maximum measure physical quantities.
The basis of measuring method of the present invention is the repeatedly average measurement of continuous sampling of high precision; In measuring process, use deque to carry out the storage of candidate's measurement data; Validity judgement and compatibility determination through measurement data are sorted out measurement data; Once reach and get rid of the disposable measurement mistake of being interrupted, allow double the measurement under the wrong situation, record correct result.Measuring process is carried out the algorithm shutdown through the maximum number of times M that measures is set under worst case, guarantee the validity and the availability of measuring method.
After measuring beginning, physical quantity is accurately measured, each measurement result is handled as follows:
One of which, measurement result at first carry out sampling queue to be handled:
(a) preceding twice measurement result directly got into sampling queue;
(b) follow-up measurement result is at first according to the error margin W that sets, and the measurement result in the sampling queue, judges the validity of this measurement data;
(c) if be judged as valid data, then directly get into sampling queue, the valid data number is added 1; If the valid data number reaches the parameter N of setting, then finish to measure, the valid data in the sampling queue are got average return;
(d) if be not sure of its validity, then judge the compatibility of data in these data and the sampling queue, as if with sampling queue in data compatibility, then upgrade the measurement data in the sampling queue;
Its two, for can't judgment data validity, with the incompatible measurement data of existing sampling queue data, carry out candidate's measurement processing again:
(a) preceding twice candidate's measurement result directly got into candidate queue;
(b) follow-up candidate's measurement result is at first according to the error margin W that sets, and data are carried out the validity judgement in the candidate queue;
(c) if be judged as valid data, then get into candidate queue, candidate's valid data number is added 1; If candidate's valid data number reaches the parameter N of setting, then finish to measure, the valid data in the candidate queue are got average return;
(d) if validity that still can not judgment data, then judge the compatibility of data in these data and the candidate queue, as if with candidate queue in data compatibility, then upgrade the measurement data in the candidate queue;
Its three, if be not sure of the validity of measurement data, measurement data is compatible with said sampling queue and candidate queue again, judges then whether the measurement number of times reaches setup parameter M:
(a) if reach setup parameter M, then finish to measure, get average according to the more queuing data of valid data in sampling queue and the candidate queue, confirm to measure the result;
(b), measure once more after then waiting for interval time of setup parameter T if do not reach setup parameter M.
The judgement of data validity can be carried out respective design according to the characteristics of measure physical quantities.Whether the mean value error that the simplest comparatively general way is checking measurements data and all available datas is in the error margin scope.As if the agreement that measured object is had span, then can add the numerical range inspection.When carrying out the data validity judgement, need to divide two kinds of situation to handle, first kind of situation is that available data validity is unknown, second kind of situation then is that available data validity is known.The processing of second kind of situation is simple relatively, only needs the error in judgement tolerance limit, span or the like.But complicacy is a bit a little in the processing of first kind of situation.Need comprehensively to judge the validity of these three data according to existing two data and the relativeness of waiting to check data.The comprehensive standard of judging of these three data validity can be set to the error of two data wherein or three data less than error margin; If the error between three data is all greater than error margin; Then get wherein adjacent two nearer data data as a reference; Be stored in formation, be not set to effectively.
Whether the judgement of data compatibility then is to detect to wait to check the magnitude relationship of data and available data compatible.If measurement data can not be judged as effectively; Be between certain two data but in the available data formation, sort; And with the error of these two data less than error margin; Then can two data that numerically are adjacent in these data and the formation be got average, merge measurement data to data queue.Do like this and can not be judged as effective the time in data, in fact effectively measurement result joins data queue, makes that the measurement result in the data queue is approached real physical quantity gradually, and transition is valid data.
The use of this method is described with the measurement example application of a reality below.Practical measuring examples is only carried out measured body weight for using electronic-weighing equipment to the pig of activity.In pig scope of activities only was limited to more among a small circle, electronic metering equipment was through the multiple spot load-bearing, and the bridge joint sensor is realized the accurate measurement to weight.Pig athletic meeting only impacts measurement result in the measuring process, and this motion has bigger random character, possibly be several motions in seconds once, also possibly be motion in tens seconds once.And the time that each motion impacts measurement result possibly continue one to two second.Therefore can produce more in the measuring process because the invalid data that the random motion incident causes, simple measuring method tends to obtain wrong measurement result.
Be an accurate Measuring Pig body weight, use this measuring method setup parameter following: measuring interval 1 second, measuring effective degree is 3 times, error margin 5%, maximum measurement number of times 10 times.The mensuration process is roughly following:
Preceding two measurement results directly get into sampling queue; Owing to have only two data, can't judge that which data is valid data, if the error between two data is less than error margin; Explain that then these two data have same alike result, promptly be all valid data or be all misdata bigger than normal.Since the 3rd data, carry out validity and judge, if be judged as effectively, then measure and finish.If can't judge validity, then carry out compatibility determination, if with sampling queue in data compatibility, then upgrade the data in the sampling queue.Otherwise measurement data will be carried out candidate queue and handled.If occur four incompatible datas continuously, then this moment, sampling queue and candidate queue all stored two measurement data.When the 5th of appearance or more incompatible datas, will directly be dropped.Current twice measurement be double random occurrence cause misdata the time, candidate queue will be carried out correct validity check and result calculating to the measurement data of back.
Claims (3)
1. fault-tolerant measuring method based on deque is characterized in that comprising the step of following order:
(1) at first, to different physical quantitys, set four parameters:
(a) the gap periods T that measures, the time of promptly repeatedly waiting between the measure physical quantities;
(b) error margin W promptly is used for confirming the error range of measurement data validity;
(c) the definite times N of measurement result is promptly confirmed the required effective measurement number of times of measurement result;
(d) the maximum number of times M that measures before promptly measurement result is confirmed, allows the number of times of maximum measure physical quantities;
(2) after the measurement beginning, physical quantity is accurately measured, each measurement result is handled as follows:
One of which, measurement result at first carry out sampling queue to be handled:
(a) preceding twice measurement result directly got into sampling queue;
(b) follow-up measurement result is at first according to the error margin W that sets, and the measurement result in the sampling queue, judges the validity of this measurement data;
(c) if be judged as valid data, then directly get into sampling queue, the valid data number is added 1; If the valid data number reaches the parameter N of setting, then finish to measure, the valid data in the sampling queue are got average return;
(d) if be not sure of its validity, then judge the compatibility of data in these data and the sampling queue, as if with sampling queue in data compatibility, then upgrade the measurement data in the sampling queue;
Its two, for can't judgment data validity, with the incompatible measurement data of existing sampling queue data, carry out candidate's measurement processing again:
(a) preceding twice candidate's measurement result directly got into candidate queue;
(b) follow-up candidate's measurement result is at first according to the error margin W that sets, and data are carried out the validity judgement in the candidate queue;
(c) if be judged as valid data, then get into candidate queue, candidate's valid data number is added 1; If candidate's valid data number reaches the parameter N of setting, then finish to measure, the valid data in the candidate queue are got average return;
(d) if validity that still can not judgment data, then judge the compatibility of data in these data and the candidate queue, as if with candidate queue in data compatibility, then upgrade the measurement data in the candidate queue;
Its three, if be not sure of the validity of measurement data, measurement data is compatible with said sampling queue and candidate queue again, judges then whether the measurement number of times reaches setup parameter M:
(a) if reach setup parameter M, then finish to measure, get average according to the more queuing data of valid data in sampling queue and the candidate queue, confirm to measure the result;
(b), measure once more after then waiting for interval time of setup parameter T if do not reach setup parameter M.
2. the fault-tolerant measuring method based on deque according to claim 1 is characterized in that: the validity judgement of measurement data is meant:
(a) when available data validity is known: whether the mean value error of measurement data and all available datas is in the error margin, if promptly effectively, otherwise is not set to effectively; Perhaps, measured object there be the regularly about of span, then add the numerical range inspection,, otherwise be not set to effective as promptly effective in span;
(b) when available data validity is unknown:, comprehensively judge the validity of these three data according to existing two data and the relativeness of waiting to check data; The comprehensive standard of judging of these three data validity is set to the error of two data wherein or three data less than error margin; If the error between three data is all greater than error margin; Then get wherein adjacent two nearer data data as a reference, be stored in formation, be not set to effectively.
3. the fault-tolerant measuring method based on deque according to claim 1 and 2; It is characterized in that: the compatibility determination of measurement data is meant: measurement data can not be judged as effectively; Be between certain two data but in the available data formation, sort; And with the error of these two data less than error margin, then be compatible.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010521783XA CN102012242B (en) | 2010-10-27 | 2010-10-27 | Fault-tolerant measurement method based on double queues |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010521783XA CN102012242B (en) | 2010-10-27 | 2010-10-27 | Fault-tolerant measurement method based on double queues |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102012242A CN102012242A (en) | 2011-04-13 |
| CN102012242B true CN102012242B (en) | 2012-05-23 |
Family
ID=43842494
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010521783XA Expired - Fee Related CN102012242B (en) | 2010-10-27 | 2010-10-27 | Fault-tolerant measurement method based on double queues |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102012242B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105700354B (en) * | 2016-01-31 | 2018-08-07 | 南通大学 | The intellegent sampling and detecting system of adjustable failure |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995033184A1 (en) * | 1994-05-26 | 1995-12-07 | Astra Aktiebolag | Measurement system and method |
| DE102004063468A1 (en) * | 2004-12-23 | 2006-07-06 | Endress + Hauser Conducta Gesellschaft für Mess- und Regeltechnik mbH + Co. KG | Method for monitoring the function of a sensor |
| CN101290611A (en) * | 2007-04-20 | 2008-10-22 | 中芯国际集成电路制造(上海)有限公司 | Data abnormal point detection method and device |
| CN101330030B (en) * | 2007-06-21 | 2010-09-29 | 中芯国际集成电路制造(上海)有限公司 | Method for removing abnormal point of detection data |
| DE102008045841A1 (en) * | 2008-09-05 | 2010-03-11 | Endress + Hauser Conducta Gesellschaft für Mess- und Regeltechnik mbH + Co. KG | Method for operating a measuring point |
-
2010
- 2010-10-27 CN CN201010521783XA patent/CN102012242B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN102012242A (en) | 2011-04-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103246602A (en) | Code coverage rate confirming method, code coverage rate confirming system, code coverage rate detecting method and code coverage rate detecting system | |
| CN109521443A (en) | A method of detection ephemeris is abnormal | |
| CN107972070A (en) | Test method, test system and the computer-readable recording medium of robot performance | |
| CN106441537A (en) | Weighing method for weighing shelves and shelf using method | |
| CN110034964A (en) | CORS data quality checking and early warning system based on TEQC | |
| CN108228722A (en) | The geographic space distribution uniformity detection method of fragmentation area sampling point | |
| CN103675022A (en) | TDS (total dissolved solids) detection method, device and system | |
| CN103164321A (en) | Occupancy rate measuring method and device of central processing unit | |
| CN111998918A (en) | Error correction method, error correction device and flow sensing system | |
| CN104142861A (en) | Processing method and processing device for configuration of server resources | |
| CN102012242B (en) | Fault-tolerant measurement method based on double queues | |
| CN110440893A (en) | Weight-calibration method, apparatus and weighing-appliance | |
| CN113172658A (en) | Robot positioning method, device, equipment and medium | |
| CN116610931B (en) | Method, device, medium and equipment for extracting numerical control countersink influencing factors of airplane | |
| CN108761161A (en) | A kind of power meter temperature influences error compensating method and device | |
| CN116703222A (en) | Method, device, electronic equipment and computer readable medium for detecting materials | |
| TWI429878B (en) | Curved surface testing system and method | |
| CN109508482A (en) | A kind of calculation method for complex-curved surface profile degree error uncertainty | |
| Delker | Evaluating risk in an abnormal world: how arbitrary probability distributions affect false accept and reject evaluation. | |
| CN102902838A (en) | Trend-based target setting method and system for process control | |
| CN105973160B (en) | Inspection method for the deformation of square hole retainer pocket hole crossbeam bend | |
| CN112765011B (en) | Quality control state judging method and device and electronic equipment | |
| CN113609971B (en) | Method, device, equipment and storage medium for checking microseism observation equipment | |
| CN108257285A (en) | A kind of data processing method of thickness transducer, device, banknote tester and storage medium | |
| CN116384168B (en) | Sand transporting amount calculating method, system, computer and storage medium of sand transporting ship |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120523 Termination date: 20151027 |
|
| EXPY | Termination of patent right or utility model |