Fusion algorithm and device for Kalman filtering of river flow based on dynamic variance

Abstract

The invention discloses a Kalman filtering fusion algorithm and device for river flow based on dynamic variance, and the method comprises the steps: (1) collecting real-time data obtained through a plurality of automatic flow monitoring methods according to different water level changes of a river; (2) converting a Kalman filtering formula into a Kalman filtering formula with multiple observation values by taking measurement results of different flow measurement instruments on the same river water body as observation values z in Kalman filtering calculation and instrument measurement errors as observation variance R; and (3) on the basis of the real-time data obtained by the various flow automatic monitoring methods, through a Kalman filtering flow fusion algorithm based on dynamic variance, obtaining fusion result data of the various flow real-time data. According to the scheme of the invention, the final calculation result does not have large jump, and the accuracy and stability of the flow measurement result are ensured.

Description

Technical Field

[0001] The present invention relates to the field of intelligent analysis of river flow measurement data, and in particular to a fusion algorithm and device for river flow based on dynamic variance Kalman filtering. Background Art

[0002] River flow is one of the important hydrological elements. Promoting full-range automatic monitoring of river flow is of great significance to flood and drought prevention, water resources management, water environment protection and water ecological restoration.

[0003] Automatic river flow monitoring methods include radar, imaging, acoustic Doppler, ultrasonic time-difference, hydraulic structure analysis, water level-discharge relationship unification, ADCP automatic flow measurement, and drop index automatic flow estimation. Most studies on automatic river flow monitoring have focused on a single method, but have not integrated multiple monitoring methods.

[0004] The closest existing technology, a river flow calculation method and system (patent number CN201910688456.4), uses radar waves to obtain historical measured flow data at the flow measurement section, including the lateral distribution data of the average flow velocity of the velocity vertical line and the large cross-sectional data of the flow measurement section; processes the lateral distribution data of the average flow velocity of the velocity vertical line, and draws a normalized standard curve of the average flow velocity of the vertical line at the measuring station; determines the outer envelope of the normalized standard curve; normalizes the current measured velocity vertical line flow data, then compares it with the normalized standard curve, corrects data that exceeds the outer envelope, and uses the corrected velocity vertical line flow data for river flow calculation. This technology uses a single radar wave measurement, which has shortcomings.

[0005] Due to the impact of climate change and human activities, as well as the limitations of the development of hydrological monitoring technology, it is difficult to realize automatic flow monitoring under conditions of low river flow and high flood conditions. Most hydrological stations find it difficult to use one method to accurately realize automatic monitoring of river flow. Different flow measuring instruments have different performances in different river water environments. Since the river water environment changes over time, the flow variance measured by each instrument will fluctuate, and the corresponding flow monitoring parameters are also dynamically adjusted. Single flow measurement methods and flow measuring instruments are not accurate, and the data differences are large. This is a problem with the current existing technology. Summary of the Invention

[0006] The present invention provides a fusion algorithm and device for river flow based on dynamic variance Kalman filtering, which can overcome the shortcomings of a single river flow measurement algorithm and obtain more accurate and reasonable flow data.

[0007] The present invention provides a fusion algorithm for river flow based on dynamic variance Kalman filtering, comprising the following steps:

[0008] (1) Collecting real-time data from various automatic flow monitoring methods in response to different water level changes in the river;

[0009] (2) Using the measurement results of the same river water body by different flow measuring instruments as the observation value z in the Kalman filter calculation, and the instrument measurement error as the observation variance R, the Kalman filter formula is transformed into a Kalman filter formula with multiple observation values;

[0010] (3) Based on the real-time data obtained by the above-mentioned multiple flow automatic monitoring methods, the fusion result data of multiple flow real-time data is obtained through the Kalman filter flow fusion algorithm based on dynamic variance.

[0011] Furthermore, the multiple automatic flow monitoring methods described in step (1) include at least two of the following: radar method, imaging method, acoustic Doppler method, ultrasonic time difference method, water level flow relationship unification method, ADCP automatic flow measurement method, and drop index automatic flow prediction method.

[0012] Furthermore, in step (2), the measurement results of the same water body using different flow measuring instruments are used as the observation value z in the Kalman filter calculation, and the instrument measurement error is used as the observation variance R; since there is more than one observation value, the Kalman filter formula is converted into the Kalman filter formula of the multi-observation value:

[0013]

[0014]

[0015] Where n is the number of observations and i is the i-th observation.

[0016] Furthermore, step (3) specifically includes:

[0017] (1) Calculate the weight w of each observation i , or the default weight of each observation value is 1 / n, where n is the number of observation values;

[0018] (2) Initialize the parameters, state covariance P0, prediction variance Q0, and initialize the flow value x'0 at the previous moment to be equal to the mean of all observations Flow change trend dx0=0;

[0019] (3) Calculate the traffic forecast value:

[0020] x k =x′ k-1 +dx k-1

[0021] (4) Calculate the observed mean:

[0022]

[0023] (5) Calculate the prediction variance:

[0024]

[0025] (6) Calculate the observation variance:

[0026]

[0027] (7) According to the five simplified Kalman filter formulas, the final fusion flow x' is calculated. k and state variance P' k ; where each observation value z i , the predicted value is x;

[0028] (8) Update the flow change trend for the next calculation:

[0029] dx k =x′ k -x′ k-1

[0030] (9) Repeat sub-steps (3) to (8) to obtain the fusion result data of multiple traffic real-time data.

[0031] Further, in step (3),

[0032] In the process of Kalman filtering algorithm, the observation error R is a fixed value. Through a variance dynamic adjustment method, the observation variance R and prediction variance Q are dynamically adjusted according to the collected data in real-time calculation, so that the full-range monitoring algorithm can adapt to the continuous changes in the water environment.

[0033] Two gradient coefficients are added to the adjustment of the prediction variance Q and the observation variance R, and the update formula becomes:

[0034] Q k =(1-α)Q k-1 +αQ′ k

[0035] R k,i =(1-β)R k-1,i +βR′ k,i

[0036] The two gradient coefficients are α and β.

[0037] Furthermore, the method further includes step (four): establishing a correlation curve with the measured flow data, calculating the correlation coefficient and system error of the correlation curve; and verifying the fusion result data.

[0038] The present invention also provides a device using the above fusion algorithm, comprising:

[0039] The data transmission module is used to collect real-time data obtained by various automatic flow monitoring instruments using corresponding monitoring methods according to different water level changes in the river through wireless or wired transmission;

[0040] The data analysis module is used to convert the Kalman filter formula into a multi-observation Kalman filter formula based on the measurement results of the same river water body by different flow measuring instruments as the observation value z in the Kalman filter calculation, and the instrument measurement error as the observation variance R;

[0041] The data fusion module is used to obtain fusion result data of multiple flow real-time data based on the real-time data obtained by the multiple flow automatic monitoring methods through a Kalman filter flow fusion algorithm based on dynamic variance.

[0042] Furthermore, the data fusion module is used to perform fusion algorithm analysis, including the following sub-steps:

[0043] (1) Calculate the weight w of each observation i , or the default weight of each observation value is 1 / n, where n is the number of observation values;

[0044] (2) Initialize the parameters, state covariance P0, prediction variance Q0, and initialize the flow value x'0 at the previous moment to be equal to the mean of all observations Flow change trend dx0=0;

[0045] (3) Calculate the traffic forecast value:

[0046] x k =x′ k-1 +dx k-1

[0047] (4) Calculate the observed mean:

[0048]

[0049] (5) Calculate the prediction variance:

[0050]

[0051] (6) Calculate the observation variance:

[0052]

[0053] (7) According to the five simplified Kalman filter formulas, the final fusion flow x' is calculated. k and state variance P' k ; where each observation value zi , the predicted value is x;

[0054] (8) Update the flow change trend for the next calculation:

[0055] dx k =x′ k -x′ k-1

[0056] (9) Repeat sub-steps (3) to (8) to obtain the fusion result data of multiple traffic real-time data.

[0057] The technical advantages of the present invention's fusion algorithm and device for Kalman filtering based on dynamic variance for river flow are as follows:

[0058] (1) Compared with the shortcomings of existing single flow measurement algorithms, the flow fusion algorithm introduced in this invention comprehensively considers the influencing factors of traditional measurement methods. Based on the Kalman filter algorithm, it transforms it into a Kalman filter formula with multiple observation values. It can dynamically adjust the variance of the predicted value and the variance of the observed value according to the data received at the current moment, and fuse to obtain a result that is more consistent with the current river flow state. The algorithm of this invention is more adaptable to the changing water environment than the traditional Kalman filter algorithm. It combines the advantages of multiple flow measurement methods to ensure that the final calculation result will not fluctuate greatly, thereby ensuring the accuracy and stability of the flow measurement results.

[0059] (2) By using the Kalman filter algorithm based on dynamic variance to fuse the flow automatic monitoring data given by multiple methods, flow automatic monitoring data with better quality than the traditional single measurement algorithm can be obtained, ensuring the rationality of the flow automatic monitoring data. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 Schematic diagram of the dynamic variance adjustment principle of the present invention.

[0061] Figure 2 Flow process diagrams obtained by various flow measurement methods.

[0062] Figure 3 This is a comparison chart of the flow measurement results of various flow measurement methods and the cruise ADCP flow measurement.

[0063] Figure 4 It is a flow process line diagram of the flow obtained by the method of the present invention and the flow obtained by the cruise ADCP.

[0064] Figure 5 This is a comparison chart of the fused flow rate obtained by the method of the present invention and the cruise-type ADCP flow measurement results. DETAILED DESCRIPTION

[0065] The technical solutions in the embodiments of the invention will be described clearly and completely below with reference to the accompanying drawings in this embodiment.

[0066] The fusion algorithm provided in the first embodiment of the present invention is a Kalman filter algorithm that can dynamically adjust the variance of each flow measured by different instruments at a hydrological station according to the current flow data, and fuse the flow to obtain a more accurate and stable flow result.

[0067] First, let's introduce the Kalman filter method. Traditional Kalman filtering is an algorithm that uses the linear system state equation and the system input and output observation data to optimally estimate the system state.

[0068] Assuming that the state prediction value is x, the state transfer matrix is F, the control input matrix is B, and the control vector is u, the state prediction formula at the kth moment is:

[0069]

[0070] Assuming the state covariance matrix is P and the prediction variance is Q, the variance update formula is:

[0071] P k =F k P k-1 F k T +Q k ②

[0072] Assuming the observation value is z, the observation variance is R, and the measurement matrix is H, the Kalman gain K is:

[0073] K=H k P k H k T (H k P k H k T +R k ) -1 =H k K′ ③

[0074] The optimal estimate of the state at the kth moment is:

[0075]

[0076] The state covariance matrix at the kth moment is updated as:

[0077] P k ′=(IK′H k )P k ⑤

[0078] The Kalman filter continuously iteratively updates the error term through the observation value and the above five formulas to minimize the error of the fusion result.

[0079] Next, the application of Kalman filtering in the flow calculation algorithm of the present invention is introduced.

[0080] The measurement results of different flow measuring instruments on the same water body are used as the observation value z in the Kalman filter calculation, and the instrument measurement error is used as the observation variance R. Since there is more than one observation value, the Kalman filter formula needs to be converted into a Kalman filter formula for multiple observation values:

[0081]

[0082] Where n is the number of observations, i is the i-th observation, and k is the k-th moment.

[0083] This research team selected the trend prediction method as the traffic prediction model, so the state transfer matrix F is the unit matrix, and the control term B k u k It is equal to the difference dx between the flow rate values at the previous moment and the moment before that. The control item generally represents the influence of external factors on the result and is a variable. However, in this patent solution, we predict that the flow rate change is a fixed change, so a constant dx is used to replace the entire control item.

[0084] Since both the observed and predicted values are flow values, the measurement matrix H is also a unit matrix. Therefore, the five formulas of the traditional Kalman filter (i.e., formulas numbered ① to ⑤) are simplified and transformed into:

[0085]

[0086] P k =P k-1 +Q k ⑦

[0087]

[0088] Next, we will introduce the dynamic variance adjustment concept of the present invention. In the traditional Kalman filter algorithm, the observation error R is a fixed value. This invention introduces a dynamic variance adjustment method. During real-time calculations, the observation variance R and the prediction variance Q are dynamically adjusted based on the collected data. This allows the algorithm to adapt to the ever-changing water environment and calculate the optimal fused flow.

[0089] Real-time calculation to obtain each observation value z i , whose mean is The variance is μ, and the predicted value is x, then theoretically, the observed mean The closer it is to the predicted value x and the smaller the observation variance μ is, the more accurate and stable the final fused traffic result will be.

[0090] See also Figure 1 , Schematic diagram of the principle of dynamic adjustment of variance. When the observed variance remains unchanged and the deviation between the observed mean and the predicted value becomes larger, we believe that the predicted value is inaccurate and we are more inclined to believe the observed value. The algorithm increases the value of the predicted variance Q; when the deviation between the observed mean and the predicted value remains unchanged and the observed variance becomes larger, we believe that the stability of the observed value is reduced and we are more inclined to believe the predicted value. The algorithm increases the value of the observed variance R.

[0091] To ensure the continuity of data changes, prevent abnormal data from causing variance mutations, and prevent large changes in Q and R that cause large fluctuations in the calculation results, the present invention adds two gradient coefficients α and β to the adjustment of the predicted variance Q and the observed variance R. The gradient coefficients range from 0 to 1. The update formula becomes:

[0092] Q k =(1-α)Q k-1 +αQ′ k

[0093] R k,i =(1-β)R k-1,i +βR′ k,i

[0094] The two gradient coefficients are α and β.

[0095] The second embodiment of the present invention provides a fusion algorithm for river flow based on dynamic variance Kalman filtering, including:

[0096] Step (1) collects real-time data from multiple automatic flow monitoring methods for different water level changes in the river. The multiple automatic flow monitoring methods include at least two of the following: radar method, imaging method, acoustic Doppler method, ultrasonic time difference method, water level flow relationship unification method, ADCP automatic flow measurement method, and drop index automatic flow estimation method.

[0097] The algorithm of the present invention collects real-time data obtained by multiple automatic flow monitoring methods, which refers to real-time data obtained by two or more existing automatic flow monitoring methods.

[0098] Step (2) The measurement results of the same river water body using different flow measurement instruments and methods are used as the observation value z in the Kalman filter calculation, and the instrument measurement error is used as the observation variance R, and the Kalman filter formula is converted into a Kalman filter formula for multiple observation values.

[0099] Specifically, since there is more than one observation value, the Kalman filter formula is transformed into the Kalman filter formula of the multiple observation values:

[0100]

[0101] Where n is the number of observations and i is the i-th observation.

[0102] Step (3) is to obtain fusion result data of multiple flow real-time data based on the real-time data obtained by the multiple flow automatic monitoring methods through a Kalman filter flow fusion algorithm based on dynamic variance.

[0103] The specific steps include the following:

[0104] (1) Calculate the weight w of each observation i , is calculated based on the jitter of the observed value change curve, or, if comparative data is available, based on the deviation between the observed value and the comparative value. If no weight is calculated, the default weight is 1 / n, where n is the number of observations.

[0105] For example, suppose that in the jth comparison test, the comparison data value is z j ', the observation value at the corresponding time is z i,j , then the root mean square error is:

[0106]

[0107] The weight is calculated as:

[0108]

[0109] (2) Initialize the parameters, state covariance P0, prediction variance Q0, and initialize the flow value x'0 at the previous moment to be equal to the mean of all observations Flow change trend dx0=0;

[0110] (3) Calculate the traffic forecast value:

[0111] x k =x′ k-1 +dx k-1

[0112] (4) Calculate the observed mean:

[0113]

[0114] (5) Calculate the prediction variance:

[0115]

[0116] (6) Calculate the observation variance:

[0117]

[0118] (7) According to the simplified Kalman filter above five formulas (i.e. formulas 11 to 12), the final fusion flow x' is calculated k and state variance P' k ; where each observation value z i , the predicted value is x;

[0119] (8) Update the flow change trend for the next calculation:

[0120] dx k =x′ k -x′ k-1

[0121] (9) Repeat sub-steps (3) to (8) to obtain the fusion result data of multiple real-time traffic data, that is, the traffic value. The traffic values obtained by different devices and methods are analyzed and integrated to obtain a more accurate and stable traffic result.

[0122] In the process of Kalman filtering algorithm, the observation error R is a fixed value. Through a variance dynamic adjustment method, the observation variance R and prediction variance Q are dynamically adjusted according to the collected data in real-time calculation, so that they can adapt to the continuous changes in the water environment.

[0123] Two gradient coefficients are added to the adjustment of the prediction variance Q and the observation variance R, and the update formula becomes:

[0124] Q k =(1-α)Q k-1 +αQ′ k

[0125] R k,i =(1-β)R k-1,i +βR′ k,i

[0126] The two gradient coefficients are α and β, and their value ranges can be adjusted between 0 and 1 according to actual conditions.

[0127] Furthermore, the method may also include step (4): establishing a correlation curve with the measured flow data, calculating the correlation coefficient and systematic error of the curve, and verifying the fusion result data. It should be noted that step (4) is optional and can be omitted, without affecting the integrity of steps (1) to (3) of the fusion algorithm of the present invention.

[0128] The advantages of the fusion algorithm of the present invention are illustrated below through comparative test results of a hydrological station.

[0129] like Figure 2Figure 3 shows flow data calculated using side-scan radar, fixed ADCP measurement, and water level drop, compared with flow measured using a propellant ADCP. This technique is used by those in this field as a standard reference, as this value best matches actual flow. Figure 2 The red dots in the middle represent the measurement results of the cruise ADCP, and the other curves of different colors represent the results of different flow measurement methods. Figure 2 The horizontal axis is the time axis.

[0130] Figure 3 In the figure, the red dashed line represents y = x, indicating that the horizontal and vertical axes are equal. The horizontal axis represents the flow rate value obtained by each measurement method, and the vertical axis represents the standard reference value (i.e., the cruise ADCP measurement result). Ideally, the two are completely equal, that is, all flow measurement points are on the red dashed line, and the ideal correlation coefficient is 1. The blue points represent the measurement values of each method. The farther they are from the red dashed line, the lower the measurement accuracy and the smaller the correlation coefficient. Figure 3 It can be seen that the measurement results of the three flow measurement methods have many abnormal values and points that deviate from the red dotted line, resulting in low accuracy of the final flow measurement value and a small correlation coefficient.

[0131] The root mean square errors (RMS) of the flow measurements obtained at the corresponding times by the underway ADCP flowmeter, the side-scan radar, the fixed ADCP, and the water level drop were calculated to be 1720.16, 773.40, and 961.73, respectively. The reciprocal values were normalized to give the weights of 0.20, 0.44, and 0.36 for each flow measurement method. The gradient coefficients α and β were set to 0.9 and 0.1, respectively.

[0132] After calculation by the fusion algorithm introduced in this invention, Figure 4 The flow rate obtained by the method of the present invention and the flow process line diagram obtained by the cruise ADCP are shown below. Figure 4 The blue curve in the middle is the flow rate obtained by the method of the present invention, and the red dots are the flow rate obtained by the cruise ADCP.

[0133] Figure 5 This is a comparison chart of the fused flow rate obtained using the method of the present invention and the flow measurement results of a cruise-type ADCP. The blue dots represent the fused flow values obtained using the method of the present invention. They are closer to the red dashed line, more consistent with the standard reference value, and have higher flow measurement accuracy and a larger correlation coefficient.

[0134] Referring to the figure above and the test comparison data table below, the correlation of the flow measurement results of the side-scan radar measurement method is 0.9916; the correlation of the flow measurement results of the fixed ADCP measurement method is 0.9986; and the correlation of the flow measurement results of the water level difference measurement method is 0.9960. The flow measurement results obtained by the fusion method of the present invention have a correlation of 0.9990, which is more consistent with the actual underway ADCP flow measurement results and has a higher correlation than any existing flow measurement method. The fused flow results obtained by the present invention have fewer outliers and jump points than any of the existing flow measurement methods, and are more accurate, reflecting that the flow measurement results of the fusion algorithm of the present invention have better accuracy and stability.

[0135] Flow measurement method name Completeness rate Correlation coefficient Systematic error Side-scan radar 80.22％ 0.9916 -3.72％ Fixed ADCP 98.71％ 0.9986 -1.07％ Water level drop measurement 97.65％ 0.9960 1.95％ Fusion Algorithm 98.92％ 0.9990 -0.82％

[0136] Test comparison data table

[0137] The dynamic variance-based Kalman filter flow fusion algorithm introduced in this paper builds on the traditional Kalman filter algorithm by adapting it to flow monitoring. It dynamically adjusts the variance of predicted and observed values based on the data received at the current moment, fusing them to produce a result that better reflects the current state. This algorithm is more adaptable to changing water environments than traditional Kalman filters, minimizing significant fluctuations in the final fused calculation results and ensuring their accuracy and stability.

[0138] Embodiment 3 of the present invention further provides a device using the above-mentioned fusion algorithm, including:

[0139] The data transmission module is used to collect and transmit real-time data obtained by various automatic flow monitoring instruments using corresponding monitoring methods according to different water level fluctuations of the river through wireless or wired transmission;

[0140] The data analysis module is used to convert the Kalman filter formula into a multi-observation Kalman filter formula based on the measurement results of the same river water body by different flow measuring instruments as the observation value z in the Kalman filter calculation, and the instrument measurement error as the observation variance R;

[0141] The data fusion module is used to obtain fusion result data of multiple flow real-time data based on the real-time data obtained by the multiple flow automatic monitoring methods through a Kalman filter flow fusion algorithm based on dynamic variance.

[0142] The specific data fusion sub-steps will not be repeated. The device using the above fusion algorithm can be a computer running the algorithm program of the present invention, or a portable monitoring device, or a system server.

Claims

1. A fusion algorithm of Kalman filter based on dynamic variance for river flow, characterized by: Including steps: (1) Collecting real-time data from various automatic flow monitoring methods in response to different water level changes in the river; (2) Using the measurement results of the same river water body by different flow measuring instruments as the observation value z in the Kalman filter calculation, and the instrument measurement error as the observation variance R, the Kalman filter formula is transformed into a Kalman filter formula with multiple observation values; (3) Based on the real-time data obtained by the above-mentioned multiple flow automatic monitoring methods, the fusion result data of multiple flow real-time data is obtained through the Kalman filter flow fusion algorithm based on dynamic variance.

2. The fusion algorithm of Kalman filter based on dynamic variance for river flow according to claim 1 is characterized in that: The multiple automatic flow monitoring methods described in step (1) include at least two of the following: radar method, imaging method, acoustic Doppler method, ultrasonic time difference method, water level flow relationship single value method, ADCP automatic flow measurement method, and drop index automatic flow prediction method.

3. The fusion algorithm of Kalman filter based on dynamic variance for river flow according to claim 1 is characterized in that: In step (2), since there is more than one observation value, the Kalman filter formula is transformed into the Kalman filter formula of the multiple observation values: Where n is the number of observations and i is the i-th observation.

4. The fusion algorithm of Kalman filter based on dynamic variance for river flow according to claim 1 is characterized in that: Step (3) specifically includes: (1) Calculate the weight w of each observation i , or the default weight of each observation value is 1 / n, where n is the number of observation values; (2) Initialize the parameters, state covariance P0, prediction variance Q0, and initialize the flow value x'0 at the previous moment to be equal to the mean of all observations Flow change trend dx0=0; (3) Calculate the traffic forecast value: x k =x′ k-1 +dx k-1 (4) Calculate the observed mean: (5) Calculate the prediction variance: (6) Calculate the observation variance: (7) According to the five simplified Kalman filter formulas, the final fusion flow x' is calculated. k and state variance P' k ; where each observation value z i , the predicted value is x; (8) Update the flow change trend for the next calculation: dx k =x′ k -x′ k-1 (9) Repeat sub-steps (3) to (8) to obtain the fusion result data of multiple traffic real-time data.

5. The fusion algorithm of Kalman filter based on dynamic variance for river flow according to claim 1 or 4 is characterized in that: In step (3), In the process of Kalman filtering algorithm, the observation error R is a fixed value. Through a variance dynamic adjustment method, the observation variance R and prediction variance Q are dynamically adjusted according to the collected data in real-time calculation, so that the full-range monitoring algorithm can adapt to the continuous changes in the water environment. Two gradient coefficients are added to the adjustment of the prediction variance Q and the observation variance R, and the update formula becomes: Q k =(1-α)Q k-1 +αQ′ k R k,i =(1-β)R k-1,i +βR′ k,i The two gradient coefficients are α and β.

6. A device using the fusion algorithm according to claim 1, characterized in that: include The data transmission module is used to collect real-time data obtained by various automatic flow monitoring instruments using corresponding monitoring methods according to different water level changes in the river through wireless or wired transmission; The data analysis module is used to convert the Kalman filter formula into a multi-observation Kalman filter formula based on the measurement results of the same river water body by different flow measuring instruments as the observation value z in the Kalman filter calculation, and the instrument measurement error as the observation variance R; The data fusion module is used to obtain fusion result data of multiple flow real-time data based on the real-time data obtained by the multiple flow automatic monitoring methods through a Kalman filter flow fusion algorithm based on dynamic variance.

7. The device according to claim 6, characterized in that The data fusion module is specifically used to perform fusion algorithm analysis, including the following sub-steps: (1) Calculate the weight w of each observation i , or the default weight of each observation value is 1 / n, where n is the number of observation values; (2) Initialize the parameters, state covariance P0, prediction variance Q0, and initialize the flow value x'0 at the previous moment to be equal to the mean of all observations Flow change trend dx0=0; (3) Calculate the traffic forecast value: x k =x′ k-1 +dx k-1 (4) Calculate the observed mean: (5) Calculate the prediction variance: (6) Calculate the observation variance: (7) According to the five simplified Kalman filter formulas, the final fusion flow x' is calculated. k and state variance P' k ; where each observation value z i , the predicted value is x; (8) Update the flow change trend for the next calculation: dx k =x′ k -x′ k-1 (9) Repeat sub-steps (3) to (8) to obtain the fusion result data of multiple traffic real-time data.

Images