Dynamic fusion measurement method based on aircraft multi-sensor time point matching

Abstract

The invention discloses a dynamic fusion measurement method based on aircraft multi-sensor time point matching, which comprises the following steps of: 1) screening measurement data to obtain credible measurement data; 2) independently grouping the credible measurement data, and determining a reference time sequence; 3) establishing a time sequence expression of credible measurement data returned by each sensor; 4) determining a time fixed domain, and constructing a time point matching model; and 5) obtaining fuel oil measurement data in the time period of the time fixed domain, and fusing the fuel oil measurement data according to the configuration condition of the sensor and the oil tank to obtain final fuel oil measurement data. The aircraft fuel quantity measurement data trend estimation method realizes aircraft fuel quantity measurement data trend estimation, enables the aircraft fuel quantity measurement data trend to be closer to a real state, eliminates oil surface fluctuation errors, overcomes measurement errors caused by multi-sensor fuel quantity data fusion during maneuvering flight, overcomes the defects that a general software filter does not consider aircraft attitude change conditions and fuel measurement data stability, and improves aircraft fuel quantity measurement accuracy. And the fuel oil measuring precision is more excellent.

Description

Technical Field

[0001] The invention relates to the technical field of liquid level measurement, and in particular to a dynamic fusion measurement method based on time point matching of multiple aircraft sensors. Background Art

[0002] At present, the fuel measurement system based on computer technology has greatly improved the accuracy of fuel measurement, but it has brought about the problem of poor measurement stability. In view of the situation of mismatched fusion of multi-source data in multi-sensor networks, it is not universal in practical applications and lacks fault tolerance. In the case of erroneous output of fuel data, it is possible to obtain erroneous fusion results, or even output fuel data that is far from the actual situation, which has seriously affected the pilot's correct judgment of the aircraft's working status. Moreover, in the process of fuel data fusion, if the waiting time is too short, it will cause too many invalid fusion times, data discreteness and low value density. If the waiting time is too long, the transmission delay will increase, which will seriously affect the reliability and effectiveness of the analysis results. Due to the uncertainty, incompleteness and instability of sensor information, the error correction ability of the fuel data fusion process is also very low, resulting in a slow convergence speed of the software filter, reduced estimation accuracy, and poor accuracy of the original data obtained, which in turn leads to suboptimal estimation of the filter or divergence of the filter, inaccurate initial value estimation, and inherent defects such as hysteresis of fuel measurement data. Summary of the invention

[0003] The purpose of the present invention is to provide a dynamic fusion measurement method based on aircraft multi-sensor time point matching, which is closer to the real state and eliminates the oil level fluctuation error.

[0004] The present invention is implemented by the following technical solution: a dynamic fusion measurement method based on aircraft multi-sensor time point matching, characterized in that it includes the following steps:

[0005] (1) Screening the fuel quantity measurement data sent back by multiple sensors, eliminating obvious false measurement data and unreliable measurement data, and obtaining reliable measurement data sent back by multiple sensors;

[0006] (2) The trusted measurement data sent back by each sensor is independently grouped and the reference time series of each group of trusted measurement data is determined;

[0007] (3) Establishing a time series expression of the credible measurement data sent back by each sensor based on the determined reference time series;

[0008] (4) Determine the time series expression to determine the time fixed domain of the trusted measurement data returned by each sensor, and then construct a time point matching model for the trusted measurement data returned by each sensor;

[0009] (5) The fuel measurement data of the time period in the fixed time domain is obtained through the time point matching model, and the fuel measurement data of multiple sensors are fused according to the configuration of the sensor and the fuel tank to obtain the final fuel measurement data.

[0010] The working principle of this technical solution is to judge the measurement data and analyze the credibility relationship between multiple sensor data under time point matching. Then, the original fuel quantity measurement data detected by each sensor is divided into groups according to the original order to determine the reference time series. By establishing the measurement data expression of each sensor equivalent to the reference time series, the time common domain of the time series of each sensor measurement data can be determined, and a multi-sensor measurement data time matching model can be constructed. Finally, the sensor measurement data fusion result of the system under the time series is solved according to the model.

[0011] In order to better implement the method of the present invention, further, in step (1), the specific process of obtaining the reliable measurement data returned by multiple sensors is as follows:

[0012] (1.1) a software filter is set to post-process the fuel quantity signal sent back by the sensor, in which a filter queue is set by a filter queue address and a filter queue length setting unit, and a filter limit, an upper limit of the filter queue, and a lower limit of the filter queue are defined;

[0013] (1.2) The filter limit is the effective range of the fuel quantity measurement data, which is a positive value. The filter limit is used to remove the fuel quantity measurement data that exceeds the filter limit as false data;

[0014] (1.3) The upper limit and the lower limit of the filter queue are taken as the credible interval, and the fuel quantity measurement data that coincides with the endpoints of the credible interval are removed as unreliable measurement data, and the fuel quantity measurement data that falls into the credible interval is retained as credible measurement data.

[0015] In order to better implement the method of the present invention, further, in step (1), the mathematical expression of the credible measurement data Y is as follows:

[0016] Y={y 1 ,...y i ,...y n},i=1,2,....n,

[0017] Where n is the number of sensors.

[0018] In order to better implement the method of the present invention, further, in step (2), the base time series expression for a period of time is as follows:

[0019] Y={y 1 ,...yi ,...y n}=y(t)

[0020] Wherein, t=0, △t, 2△t, ... i△t, ... (n-1)△t, i is the ith sensor, △t is the sensor measurement data interval, and y(t) is the fuel quantity measurement data of the sensor at time t.

[0021] In order to better implement the method of the present invention, further, in the step (3), the time series expression of the trusted measurement data sent back by each sensor is established as follows:

[0022] f′ i (t) = f i (t+Z i Δt)+k i,j (Δt i -Z i Δt)

[0023] Among them, f i `(t) represents the time series expression of the fuel quantity measurement number of each sensor i under the reference time series, K i，j It represents the slope of the two fuel quantity measurement data of the i-th sensor at the j-th moment under the reference time, j represents the measurement moment number, j=1,2,3...N, N is the number of measurement time periods, and the slope here represents the positive or negative relationship between the fuel quantity measurement data of the two sensors, as well as the speed of change. If the slope is positive, there is a positive relationship, indicating that the two fuel quantity measurement data numbers increase or decrease at the same time; if the slope is negative, it means that the two fuel quantity measurement data have an inverse relationship, that is, when one fuel quantity measurement number increases, the other fuel quantity measurement data decreases.

[0024] In order to better implement the method of the present invention, further, in the step (3), before establishing the time series expression of the reliable measurement data returned by each sensor, using △t i and △t represent the number of measurement intervals between the time series of the sensor fuel quantity measurement data and the time series of the reference time series, recorded as Z i , which is expressed as follows:

[0025]

[0026] Among them, △t i It represents the difference between the time series of the fuel quantity measurement data of the i-th sensor and the reference time, and △t is the interval of the fuel quantity measurement data of the sensor;

[0027] Then, the slope of the line connecting the two fuel quantity measurement data of the i-th sensor at the j-th moment in the reference time series is expressed by combining the fuel quantity measurement data of the two adjacent sensors, and is recorded as K i，j , which is expressed as follows:

[0028]

[0029] K i，j It represents the slope of the two fuel quantity measurement data of the i-th sensor at the j-th moment under the reference time, j represents the measurement moment number, j=1,2,3...N, N is the number of measurement time periods.

[0030] In order to better implement the method of the present invention, further, the time fixed domain determination process in step (4) is as follows: let the difference between the time series of n sensor measurement data and the reference time series be expressed as {△t 1 , △t 2 , ... △t, ... △t n The minimum and maximum values ​​of the difference between the time series of each sensor's measured data and the reference time series are shown in the following formula:

[0031]

[0032] Among them, △t imin , △t imax Respectively represent the minimum and maximum values ​​of each sensor measurement data time series relative to the reference time series;

[0033] Thus, the first position Z of the time-fixed domain is obtained imin With tail position Z imax , so the expression of the time fixed domain is as follows:

[0034] t=0,…,iΔt,…,(n-2-Z imax )Δt,Δt imin ≥0

[0035] t=(|Z imin |+1)Δt,…,iΔt,…,(n-1)Δt,Δt imax ≤0

[0036] t=(|Z imin |+1)Δt,…,iΔt,…,(n-2-|Z imax |)Δt,Δt imin <0,Δt imax >0

[0037] Where t represents the time fixed domain of the time series of each sensor measurement data; △t imin , △t imaxRespectively represent the minimum and maximum values ​​of the difference between the time series of each sensor measurement data and the reference time series; Z imin With Z imax Respectively represent the minimum and maximum values ​​of the number of measurement data intervals that differ between the time of each sensor's measurement data time series and the reference time series time.

[0038] In order to better implement the method of the present invention, further, in the step (4), a time point matching model of the reliable measurement data returned by each sensor is constructed as follows:

[0039]

[0040] Among them, y(t) represents the total oil volume in the system during this period.

[0041] In order to better implement the method of the present invention, further, in the step (5), the configuration of the sensor and the fuel tank includes two situations, one of which is to test the fuel amount in only one fuel tank, and the fuel tank is provided with n sensors, then the fuel measurement data of multiple sensors are fused, and the final fuel measurement data is calculated according to the following formula:

[0042]

[0043] The other method is to test the fuel quantity of n fuel tanks. Only one sensor is set in each fuel tank. Then the fuel measurement data of multiple sensors are fused, and the final fuel measurement data is calculated according to the following formula:

[0044]

[0045] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0046] The present invention realizes the trend estimation of aircraft fuel quantity measurement data, making it closer to the actual state, eliminating the error of oil level fluctuation, overcoming the measurement error caused by the fusion of multi-sensor fuel quantity data during maneuvering flight, and effectively improving the technical defects of general software filters that do not consider the situation of aircraft attitude changes and the stability of fuel measurement data, so that the measurement method has more superior fuel measurement accuracy. DETAILED DESCRIPTION

[0047] In order to make the purpose, process conditions and advantages of the present invention more clearly understood, the present invention is further described in detail in conjunction with the following embodiments, but the embodiments of the present invention are not limited thereto. Without departing from the above-mentioned technical ideas of the present invention, various substitutions and changes are made according to the common technical knowledge and customary means in the field, which should all be included in the scope of the present invention. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0048] Embodiment 1:

[0049] This embodiment provides a dynamic fusion measurement method based on aircraft multi-sensor time point matching, which is as follows:

[0050] In the software filter, a filter queue address, a confidence interval length setting unit, a filter limit, a filter queue upper limit, and a filter queue lower limit are set to obtain reliable measurement data. The filter queue address defines the storage starting position of the measurement data, which is convenient for us to find the initial value of the filter. The confidence interval length setting unit defines the response speed and interval length of the filter, which is convenient for us to find the number of data that need to be filtered. The filter limit defines the value range of the measurement data to eliminate obviously false data. The filter queue upper limit and the filter queue lower limit define the confidence interval range of the filter sequence to distinguish the reliability of the measurement data. The reliable measurement data defines the output interface of the reliable measurement data. Its working principle is to first use the filter limit Δ limit Remove obvious false data; then use the upper and lower limits of the filter queue Δ upper and Δ lower The credibility of the measured data is distinguished to obtain the credible measured data; finally, the software filter outputs the credible measured data. limit S is the oil quantity measurement data i The effective range is positive, Δ upper and Δ lower They are the upper and lower limits of the filter queue respectively.

[0051] Since the actual fuel consumption of the aircraft is reduced at a relatively stable fuel consumption rate, the concept of reliable measurement data of fuel quantity is introduced here: reliable measurement data x i When it changes downward within a certain range, the credible measurement data x i The minimum relative base oil volume is not less than Δ lower , can also change upward, but the maximum cannot exceed Δ upper .

[0052] After the software filter starts running, the measured data S i , it is necessary to fill the data queue after n measurement cycles as the basis for subsequent comparison. On the basis of the queue being filled, firstly i With Δ limit Comparison can exclude obvious false data; sort according to the original order, if S i Less than Δ limit And in [Δ lower ,Δ upper ] interval, it is considered to be credible and enters the confidence interval.

[0053] The mathematical expression of the reliable measurement data Y of the software filter is:

[0054] Y={y 1 ,...y i ,...y n},i=1,2,....n (1)

[0055] A time series is an ordered set of elements consisting of the recorded values ​​of a physical quantity and time nodes. In a generalized time series, the recorded values ​​are not limited to numerical values, but can be multimedia data, discrete symbols, custom model data, and other information. In a narrow time series, the recorded values ​​generally refer to a real value. Then the benchmark time series over a period of time can be expressed as:

[0056]

[0057] Wherein, t=0,Δt,2Δt,...,iΔt,...,(n-1)Δt; y(t) represents the fuel data measured at time t; Δt represents the interval time of the measurement data.

[0058] If the measured data falls within the confidence interval condition, the measured data benchmark time series is obtained using formula (2);

[0059] Before establishing the expression of each sensor measurement data time series relative to the reference time series, use Δt i and Δt represent the number of measurement intervals between the sensor measurement data time series and the reference time series time, denoted as Z i , according to the rounding function, its representation is as follows:

[0060]

[0061] Among them, Δt i It represents the difference between the time series time of the i-th sensor measurement data and the reference time, and Δt represents the measurement data interval time.

[0062] Then the slope of the line connecting the two measured data of the i-th sensor at the j-th moment under the reference time series represented by the two adjacent sensor data is recorded as k i,j , which is expressed as follows:

[0063]

[0064] Among them, k i,j It represents the slope of two measurement data of the i-th sensor at the j-th time under the reference time, j represents the measurement time number. j=1,2,3,,...N, N represents the number of measurement time periods.

[0065] According to formulas (3)-(6), the expression of each sensor measurement data time series relative to the reference time series is as follows:

[0066] f′ i (t) = f i (t+Z i Δt)+k i,j (Δt i -Z i Δt) (7)

[0067] Among them, f′ i (t) represents the time series expression of the measurement data of each sensor in the reference time series.

[0068] According to formulas (3)-(7), expressions of the time series of each sensor measurement data relative to the reference time series are established. Formula (7) is the expression of the time series of each sensor measurement data relative to the reference time series.

[0069] In order to ensure that the time series of sensor measurement data with time point matching are fused within a fixed time region, the concept of time fixed domain is introduced here, that is, the time series of sensor measurement data with time point matching have a fixed time region.

[0070] Assume that the difference between the time series of n sensor measurement data and the reference time series is expressed as: {Δt 1 ,Δt 2 ,...,Δt,...,Δt n}, take the minimum and maximum difference between the time series of each sensor measurement data and the reference time series as shown in the following formula:

[0071]

[0072] Among them, Δt imin , Δt imax They respectively represent the minimum and maximum values ​​of each sensor measurement data time series relative to the reference time series.

[0073] Δt imin , Δt imax Substituting into formula (3) and (4), we can obtain Z imin , Z imax According to Z imin , Z imax Determine the first and last positions of the time fixed domain, which is expressed as follows:

[0074] t=0,…,iΔt,…,(n-2-Z imax )Δt,Δt imin ≥0 (10)

[0075] t=(|Z imin |+1)Δt,…,iΔt,…,(n-1)Δt,Δt imax ≤0 (11)

[0076] t=(|Z imin |+1)Δt,…,iΔt,…,(n-2-|Z imax |)Δt,Δt imin <0,Δt imax >0 (12)

[0077] Where t represents the fixed time domain of the time series of each sensor measurement data; Δt imin , Δt imax Respectively represent the minimum and maximum values ​​of the difference between the time series of each sensor measurement data and the reference time series; Z imin , Z imax Respectively represent the minimum and maximum values ​​of the number of measurement data intervals that differ between the time of each sensor's measurement data time series and the reference time series time.

[0078] Then, the time point matching model is established as follows:

[0079]

[0080] y(t) represents the total oil volume of the system during this period. Formula (13) is equal to the total oil volume of the system, which is obtained by accumulating the measurement data of each sensor in the time series. Formulas (8) and (9) are to clarify the Δt range of each i-th sensor, and formulas (10)-(12) are to clarify the time series t range of each i-th sensor, f i () is the expression of the i-th sensor in time series.

[0081] Finally, combining all the above formulas, we get the time point matching dynamic fusion calculation formula (14):

[0082] The calculation formula (14) is

[0083] Finally, based on the input measurement data and time interval, the model algorithm can be used to iteratively obtain the results of multi-sensor measurement data fusion.

[0084] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A dynamic fusion measurement method based on aircraft multi-sensor time point matching, characterized in that: The following steps are involved: (1) Screening the fuel quantity measurement data sent back by multiple sensors, eliminating obvious false measurement data and unreliable measurement data, and obtaining reliable measurement data sent back by multiple sensors; (2) The trusted measurement data sent back by each sensor is independently grouped and the reference time series of each group of trusted measurement data is determined; (3) Establishing a time series expression of the credible measurement data sent back by each sensor based on the determined reference time series; (4) Determine the time series expression to determine the time fixed domain of the trusted measurement data returned by each sensor, and then construct a time point matching model for the trusted measurement data returned by each sensor; (5) The fuel measurement data of the time period in the fixed time domain is obtained through the time point matching model, and the fuel measurement data of multiple sensors are fused according to the configuration of the sensor and the fuel tank to obtain the final fuel measurement data.

2. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to claim 1 is characterized in that: In step (1), the specific process of obtaining the reliable measurement data sent back by multiple sensors is as follows: (1.1) a software filter is set to post-process the fuel quantity signal sent back by the sensor, in which a filter queue is set by a filter queue address and a filter queue length setting unit, and a filter limit, an upper limit of the filter queue, and a lower limit of the filter queue are defined; (1.2) The filter limit is the effective range of the fuel quantity measurement data, which is a positive value. The filter limit is used to remove the fuel quantity measurement data that exceeds the filter limit as false data; (1.3) The upper limit and the lower limit of the filter queue are taken as the credible interval, and the fuel quantity measurement data that coincides with the endpoints of the credible interval are removed as unreliable measurement data, and the fuel quantity measurement data that falls into the credible interval is retained as credible measurement data.

3. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to claim 2 is characterized in that: In step (1), the mathematical expression of the credible measurement data Y is as follows: Y={y1,...y i ,...y n },i=1,2,....n, Where n is the number of sensors.

4. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to any one of claims 1 to 3, characterized in that: In step (2), the base time series expression for a period of time is as follows: Y={y1,...y i ,...y n }=y(t) Wherein, t=0, △t, 2△t, ... i△t, ... (n-1)△t, i is the ith sensor, △t is the sensor measurement data interval, and y(t) is the fuel quantity measurement data of the sensor at time t.

5. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to any one of claims 1 to 3, characterized in that: In step (3), the time series expression of the reliable measurement data sent back by each sensor is established as follows: f i ′(t)=f i (t+Z i Δt)+k i,j (Δt i -Z i Δt) Among them, f i `(t) represents the time series expression of the fuel quantity measurement number of each sensor i under the reference time series, K i，j It represents the slope of the two fuel quantity measurement data of the i-th sensor at the j-th moment under the reference time, j represents the measurement moment number, j=1,2,3...N, N is the number of measurement time periods, and the slope here represents the positive or negative relationship between the fuel quantity measurement data of the two sensors, as well as the speed of change. If the slope is positive, there is a positive relationship, indicating that the two fuel quantity measurement data numbers increase or decrease at the same time; if the slope is negative, it means that the two fuel quantity measurement data have an inverse relationship, that is, when one fuel quantity measurement number increases, the other fuel quantity measurement data decreases.

6. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to claim 5 is characterized in that: In step (3), before establishing the time series expression of the reliable measurement data sent back by each sensor, △t i and △t represent the number of measurement intervals between the time series of the sensor fuel quantity measurement data and the time series of the reference time series, recorded as Z i , which is expressed as follows: Among them, △t i It represents the difference between the time series of the fuel quantity measurement data of the i-th sensor and the reference time, and △t is the interval of the fuel quantity measurement data of the sensor; Then, the slope of the line connecting the two fuel quantity measurement data of the i-th sensor at the j-th moment in the reference time series is expressed by combining the fuel quantity measurement data of the two adjacent sensors, and is recorded as K i，j , which is expressed as follows: K i，j It represents the slope of the two fuel quantity measurement data of the i-th sensor at the j-th moment under the reference time, j represents the measurement moment number, j=1,2,3...N, N is the number of measurement time periods.

7. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to any one of claims 1 to 3, characterized in that: The process of determining the time fixed domain in step (4) is as follows: let the difference between the time series of n sensor measurement data and the reference time series be expressed as {△t1, △t2, ... △t, ... △t n The minimum and maximum values ​​of the difference between the time series of each sensor's measured data and the reference time series are shown in the following formula: Among them, △t imin , △t imax Respectively represent the minimum and maximum values ​​of each sensor measurement data time series relative to the reference time series; Thus, the first position Z of the time-fixed domain is obtained imin With tail position Z imax , so the expression of the time fixed domain is as follows: t=0,…,iΔt,…,(n-2-Z i max )Δt,Δt i min ≥0 t=(|Z imin |+1)Δt,…,iΔt,...,(n-1)Δt,Δt i max ≤0 t=(|Z i min |+1)Δt,...,iΔt,...,(n-2-|Z i max |)Δt,Δt i min <0,Δt i max >0 Where t represents the time fixed domain of the time series of each sensor measurement data; △t imin , △t imax Respectively represent the minimum and maximum values ​​of the difference between the time series of each sensor measurement data and the reference time series; Z imin With Z imax Respectively represent the minimum and maximum values ​​of the number of measurement data intervals that differ between the time of each sensor's measurement data time series and the reference time series time.

8. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to claim 7 is characterized in that: In step (4), a time point matching model of the reliable measurement data returned by each sensor is constructed as follows: Among them, y(t) represents the total oil volume in the system during this period.

9. The method for dynamic fusion measurement based on aircraft multi-sensor time point matching according to claim 7 is characterized in that: In step (5), there are two configurations of the sensor and the fuel tank. One of them is to test the fuel level in only one fuel tank. The fuel tank is provided with n sensors, and the fuel measurement data of multiple sensors are fused. The final fuel measurement data is calculated according to the following formula: The other method is to test the fuel quantity of n fuel tanks. Only one sensor is set in each fuel tank. Then the fuel measurement data of multiple sensors are fused, and the final fuel measurement data is calculated according to the following formula:

Images