Method for improving initial pointing precision of laser communication terminal

Abstract

The invention relates to a method for improving initial pointing precision of a laser communication terminal. The method comprises the following steps that an equipment terminal is initialized to operate, and a double-end terminal turntable is adjusted according to positioning data output by a GPS positioning system; obtaining attitude data of the rotary table through the master control slave equipment terminal; performing mean filtering operation on the data in the process of acquiring the data to suppress high-frequency interference components in the data; solving a fitting value of the group of data by utilizing a linear fitting method; solving a low-frequency interference component in the group of data through the fitting value and a data result obtained through operation in the step 2; calculating attitude data after the low-frequency interference component is removed, wherein the data is used as an attitude measurement value after the low-frequency interference component is removed; and performing Kalman filtering operation, taking a correction value obtained by the set operation as a pointing attitude value of the turntable, and fixing the terminal turntable according to the obtained attitude value of the turntable. According to the method, the problem of aiming direction deviation of the laser communication terminal equipment caused by data errors generated by GPS interference can be effectively solved.

Description

Technical Field

[0001] The invention relates to a laser communication terminal calculation method, in particular to a method for improving the initial pointing accuracy of a laser communication terminal. Background Art

[0002] In the field of wireless laser communication, the workflow of terminal equipment is mainly divided into three stages: aiming, capturing, and tracking. In the aiming stage, the dual-end equipment will adjust the attitude of the terminal turntable in real time according to the positioning data given by GPS to perform initial pointing so that the dual-end equipment points accurately. However, after the pointing is completed, the dual-end turntable will rotate randomly to varying degrees. The main reason is that the GPS positioning system is affected by various uncertain factors in this process, such as weather conditions, troposphere, ionosphere and other interferences. These influencing factors will cause the positioning data output by GPS to change continuously, thereby causing the device terminal turntable to rotate randomly. Only when the terminal device enters the next working state can the terminal device stop acquiring GPS data and stop rotating. This will cause the attitude angle of the turntable to have random errors when it stops rotating before the terminal device enters the capture stage, which will eventually cause deviations in the aiming and pointing. Summary of the invention

[0003] The purpose of the present invention is to overcome the above shortcomings and provide a method for improving the initial pointing accuracy of the laser communication terminal, which can effectively solve the problem of aiming and pointing deviation of the laser communication terminal equipment caused by data errors caused by GPS interference. This improves the pointing accuracy of the laser communication terminal to a certain extent, laying a good foundation for the subsequent second-stage capture.

[0004] According to the technical solution provided by the present invention, a method for improving the initial pointing accuracy of a laser communication terminal comprises the following steps:

[0005] Step 1: The equipment terminal is initialized and runs, and the dual-end terminal turntable is adjusted according to the positioning data output by the GPS positioning system, so that the dual-end equipment can be initially aligned;

[0006] Step 2, after step 1 is executed, the attitude data of the turntable is obtained from the device terminal through the main control, and the attitude data includes the azimuth data and pitch data of the terminal turntable; in the process of obtaining the data, the data is subjected to mean filtering operation to suppress the high-frequency interference components in the data;

[0007] Step 3, according to the data result obtained by the operation in step 2, the fitting value of the group of data is obtained by using the linear fitting method;

[0008] Step 4, according to the fitting value obtained in step 3, the component of low-frequency interference in the group of data is calculated by combining the fitting value with the data result obtained by the operation in step 2;

[0009] Step 5, according to the posture data obtained in step 2 and the low-frequency interference component in the data obtained in step 4, calculate the posture data after removing the low-frequency interference component, and use the data as the posture measurement value after removing the low-frequency component;

[0010] Step 6, based on the processed measurement value obtained in step 5, perform Kalman filter operation on it. When the number of operations reaches the set number, take the correction value obtained by the current operation as the pointing attitude value of the turntable, and fix the terminal turntable through the main control according to the obtained turntable attitude value.

[0011] As a further improvement of the present invention, the mean filtering operation in step 2 is specifically as follows:

[0012] 3) Take 150 azimuth or elevation raw data, the sliding window size is 1*10, and the step size is 1. In this way, the amount of data of the calculation result is only 141. In order to keep the amount of data of the calculation result unchanged, 9 data need to be added;

[0013] 4) Take 150 azimuth data, each group of data is: θ fw(1) ~θ fw(10) ,θ fw(2) ~θ fw(11) ,θ fw(3) ~θ fw(12) Until θ fw(150) ~θ fw(159) ; For each set of posture data (θ fw(j) ,θ fy(j) ) are respectively processed by mean filtering, where θ fw is the position data of the turntable, θ fy is the pitch data of the turntable, and the mean value of each group is: (θ fwc(i) ,θ fyc(i) ), the calculation formula is:

[0014]

[0015] Where i is the group for mean filtering operation, i∈(0,149), j is the serial number of the supplemented data, j∈(1,159), θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

[0016] As a further improvement of the present invention, the method of supplementing data is: every 15 original data are supplemented with the mean of these 15 data, and no supplement is needed after the last 15 data, so that 9 data will be added to the 150 original data, and the amount of data after supplementation becomes 159. At this time, the above-mentioned mean filtering operation is performed on it, and the resulting data amount becomes 150, and the data amount is the same as the original data amount before supplementation.

[0017] As a further improvement of the present invention, the fitting value (θ fwc ,θ fyc ) is calculated as:

[0018]

[0019] where θ fwc is the fitting value of the data after the mean filtering of the azimuth data, θ fyc is the fitting value of the pitch data after mean filtering, θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

[0020] As a further improvement of the present invention, the low-frequency interference component in the data in step 4: (θ fwcz(i) ,θ fycz(i) ), which is calculated as follows:

[0021] θ fwcz(i) =θ fwc(i) -θ fwc Formula (5)

[0022] θ fycz(i) =θ fyc(i) -θ fyc Formula (6)

[0023] where θ fwcz(i) is the low-frequency interference component in the azimuth data, θ fycz(i) is the low-frequency interference component in the pitch data, θ fwc is the fitting value of the data after the mean filtering of the azimuth data, θ fyc is the fitting value of the pitch data after mean filtering, θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

[0024] As a further improvement of the present invention, the posture measurement value calculation method is: according to the original data (θ fwor(t) ,θ fyor(t) ), t is the serial number of the original posture data obtained by the master control, t∈(1,150); the low-frequency interference component (θ fwcz(t-1) ,θ fycz(t-1) ), and obtain the processed measurement value (θ fwme(t) ,θ fyme(t) ), the calculation formula is as follows:

[0025] θ fwme(t) =θ fwor(t)- θfwcz(t-1) Formula (7)

[0026] θ fyme(t) =θ fyor(t)- θ fycz(t-1) Formula (8)

[0027] where θ fwme(t) is the measured value of the orientation after processing, θ fyme(t) is the measured value of pitch after processing, θ fwor(t) is the original orientation data, θ fyor(t) is the original data of pitch; θ fwcz(t-1) is the low-frequency interference component in the azimuth data, θ fycz(t-1) is the low-frequency interference component in the pitch data.

[0028] As a further improvement of the present invention, the formula for the Kalman filter operation is:

[0029]

[0030] x k =x (k-1) +k k (z k -x (k-1) ) Formula (10)

[0031] e EST(k) =(1-k k ) EST(k-1) Formula (11)

[0032] e EST(k-1) is the last correction error, x (k-1) is the last correction value, z k is the current measured value, k k is the current Kalman gain constant, x (k) is the current correction value, e EST(k) is the current correction error.

[0033] As a further improvement of the present invention, before the Kalman filter operation, an initial correction error e is given according to the actual performance of the device. EST(0) , initial correction value x0 and measurement error e MEA(k) , the calculation process is divided into three steps:

[0034] 1) Calculate the current Kalman gain k k ;

[0035] 2) Calculate the current correction value x k ;

[0036] 3) Finally calculate the current correction error e EST(k) ;

[0037] A total of 150 posture data are collected through step 5, and all the data are sequentially substituted into formulas (9) to (11) to update the gain k k and the corrected error e EST(k) When the number of operations reaches the set number, the current correction value is taken as the pointing attitude value of the terminal turntable, and the terminal turntable is fixed according to the obtained attitude value through the main control.

[0038] The beneficial effects of the present invention are:

[0039] The present invention fully considers that the equipment terminal turntable is affected by GPS data errors, and the random rotation of the turntable has normal distribution. The Kalman filter, which is an algorithm with good suppression of normal distribution noise, is used to better suppress the influence of GPS signal disturbance on the terminal turntable.

[0040] At the same time, taking into account the actual situation at the working site of the equipment terminal, the equipment terminal will be affected by wind forces to varying degrees in different time periods. In the data preprocessing stage, the present invention adopts a combination of mean filtering and linear fitting to suppress the low-frequency disturbance part of the data. This method reduces the impact of wind disturbances and lays a good foundation for the execution of subsequent algorithms.

[0041] Through the above processing method, the present invention suppresses the wind disturbance at the terminal equipment working site, and at the same time better suppresses the random error caused by GPS signal disturbance, thereby more efficiently improving the initial pointing accuracy of the laser communication terminal, laying a good foundation for the subsequent capture work of the device terminal. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Specific operation flow chart.

[0043] Figure 2 Comparison chart of supplemented data and original data.

[0044] Figure 3 Distribution diagram of mean filtering and linear fitting results.

[0045] Figure 4 Graph of low-frequency component calculation results.

[0046] Figure 5 Comparison of the results before and after removing the low-frequency components of the original data.

[0047] Figure 6 Kalman filter operation process diagram.

[0048] Figure 7 Comparison of the results of two Kalman filtering methods.

[0049] Description of reference numerals: Figure 2 ,3 , 4, 5, and 7, the numerical units of the vertical coordinates are all angle units (degrees), and the horizontal coordinates are all serial numbers that are constants. Figure 6 Except for the Kalman gain which is a constant, all other numerical units are angle units (degrees). DETAILED DESCRIPTION

[0050] The present invention will be further described below with reference to the embodiments in the accompanying drawings:

[0051] As shown in the figure, a method for improving the initial pointing accuracy of a laser communication terminal includes the following steps:

[0052] Step 1: The device is powered on and enters the initialization running state. After the dual-end device enters the running state, the posture is adjusted according to the positioning data transmitted by the GPS, so that the dual-end devices are initially aligned, and the device terminal enters the aiming working state.

[0053] Step 2. After step 1 is executed, the main control obtains attitude data from the device terminal. The attitude data includes the azimuth data and pitch data of the device terminal turntable. During the acquisition process, they are respectively subjected to mean filtering, where the sliding window size is 1*10 and the step length is 1. In this way, the data volume of the calculation result is only 141. In order to keep the data volume of the calculation result unchanged, the original data needs to be supplemented. The supplementation method is: every 15 original data are supplemented with the mean of these 15 data, and the last 15 data do not need to be supplemented. In this way, 9 data will be added to the 150 original data. The amount of data after supplementation becomes 159. At this time, the above-mentioned mean filtering operation is performed on it, and the amount of data of the result becomes 150. At this time, the amount of data is the same as the original data before supplementation. At this time, the mean filtering operation is performed on the supplemented data. Taking the azimuth data as an example, the data groups are: θ fw(1) ~θ fw(10) ,θ fw(2) ~θ fw(11) ,θ fw(3) ~θ fw(12) Until θ fw(150) ~θ fw(159) Then for each set of posture data (θ fw(j) ,θ fy(j) ) are respectively subjected to mean filtering, among which θ fw is the position data of the turntable, θ fy is the pitch data of the turntable. The present invention uses the mean filtering algorithm to suppress the high-frequency components in the data, and the mean result of each group is: (θ fwc(i) ,θ fyc(i) ), the calculation formula is:

[0054]

[0055] Where i is the group for mean filtering operation, i∈(0,149), j is the serial number of the supplemented data, j∈(1,159), θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

[0056] Step 3: According to the posture data after mean filtering obtained in step 2, the high-frequency interference component in the data has been suppressed. At this time, the fitting value of the mean filtered data is obtained by linear fitting: (θ fwc ,θ fyc ), the calculation formula is:

[0057]

[0058] where θ fwc is the fitting value of the data after the mean filtering of the azimuth data, θ fyc is the fitting value of the pitch data after mean filtering, θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

[0059] Step 4: Based on the fitting value obtained in step 3, the component of low-frequency interference in the data set is obtained by combining the fitting value with the data result obtained in step 2: (θ fwcz(i) ,θ fycz(i) ), which is calculated as follows:

[0060] θ fwcz(i) =θ fwc(i) -θ fwc Formula (5)

[0061] θ fycz(i) =θ fyc(i) -θ fyc Formula (6)

[0062] where θ fwcz(i) is the low-frequency interference component in the azimuth data, θ fycz(i) is the low-frequency interference component in the pitch data, θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data, θ fwc is the fitting value of the data after the mean filtering of the azimuth data, θ fyc is the fitted value of the pitch data after mean filtering.

[0063] Step 5, according to the original data obtained in step 2 (excluding supplementary data): (θ fwor( t),θ fyor(t)), t is the serial number of the original posture data obtained by the master control, t∈(1,150). Subtract the low-frequency interference component (θ fwcz(t-1) ,θ fycz(t-1) ), and obtain the processed measurement value (θ fwme(t) ,θ fyme(t) ), the calculation formula is as follows:

[0064] θ fwme(t) =θ fwor(t) -θ fwcz(t-1) Formula (7)

[0065] θ fyme(t) =θ fyor(t)- θ fycz(t-1) Formula (8)

[0066] where θ fwme(t) is the measured value of the orientation after processing, θ fyme(t) is the measured value of pitch after processing, θ fwor(t) is the original orientation data, θ fyor(t) is the original data of pitch; θ fwcz(t-1) is the low-frequency interference component in the azimuth data, θ fycz(t-1) is the low-frequency interference component in the pitch data.

[0067] Step 6: Perform Kalman filter operation on the processed measurement value obtained in step 5. When the number of operations reaches the set number, the current correction value is taken as the final pointing attitude value of the turntable. The calculation formula is as follows:

[0068]

[0069] x k =x (k-1) +k k (z k -x (k-1) ) Formula (10)

[0070] e EST(k) =(1-k k ) EST(k-1) Formula (11)

[0071] As described in the above formula, the initial correction error e must be given according to the actual performance of the equipment before calculation. EST(0) , initial correction value x0 and measurement error e MEA(k) , k k is the current Kalman gain constant, x (k) is the current correction value, e EST(k) is the current correction error. The calculation process is divided into three steps. First, the current Kalman gain k is calculated. k, then calculate the current correction value x k , and finally calculate the current correction error e EST(k) , e EST(k-1) is the last correction error, x (k-1) is the last correction value, z k The present invention collects 150 posture data in total through step 5, and puts all the data into formulas (9) to (11) in turn to update the gain k k and the corrected error e EST(k) When the number of operations reaches the set number, the current correction value is taken as the pointing attitude value of the terminal turntable, and the terminal turntable is fixed according to the obtained attitude value through the main control.

[0072] Example

[0073] The present invention is effective for fixing the double-ended turntable of laser communication in the aiming state. This example is only illustrated by obtaining the posture data of a device at one end. This time, 150 turntable posture data of a device at one end after the first step is executed are recorded. Only part of the orientation data in the posture data is processed for illustration, and the pitch data processing method is the same.

[0074] After the device terminal is powered on and enters the initialization running state, the dual-end devices adjust their posture according to the positioning information transmitted by the GPS positioning system. At this time, the dual-end devices achieve preliminary alignment. At this time, the posture data of one end device is queried through the main control, and the alignment is supplemented during the query process. The supplement method is: every time 15 posture data are queried, the average value of these 15 data is supplemented afterwards. For the 150 data queried this time, 9 values ​​need to be supplemented, and the last 15 values ​​queried do not need to be supplemented. The distribution of the original data and the supplemented data is as follows Figure 2 As shown, the blue curve is the distribution of the original data, and the red curve is the distribution of the original data after supplementation.

[0075] Next, the data supplemented in step 2 needs to be processed by mean filtering using formula (1), with a sliding window size of 1*10 and a step size of 1. The data obtained by mean filtering is then fitted. The fitting value obtained by formula (3) is: -88.774. The result is as follows: Figure 3 As shown, the blue curve is the distribution of the original data after supplementation, and the red curve is the data distribution after mean filtering of the supplemented data. It can be seen that the high-frequency noise of the data is well suppressed after mean filtering, and the ordinate of the green straight line is the fitting value of the data after mean filtering.

[0076] According to the fitting value data obtained in step 3, the corresponding difference is obtained by subtracting the fitting value of the mean filtered data from the mean filtered data according to formula (5). The difference is regarded as the low-frequency component in the azimuth data. The result is as follows: Figure 4 shown.

[0077] Then, according to the original data obtained in step 2 (excluding the supplementary data), the difference obtained in step 4 is subtracted from it according to formula (7) to obtain the processed measurement value. The result is as follows: Figure 5 As shown, the blue curve is the original data, and the yellow curve is the measured value obtained after processing in the previous steps. It can be seen that the low-frequency disturbance part of the processed data is well suppressed.

[0078] Finally, the processed measurement values ​​obtained in step 5 are filtered using the Kalman filter algorithm. First, according to the measured data analysis, it is concluded that the range of the azimuth attitude data is generally stable within 4, so the measurement error is located at 2. Then, according to the turntable at one end of the observation, the azimuth value is set to -90, and the initial estimation error is set to 3. Taking the four processed measurement values ​​obtained in step 5 as an example, they are: -89.15148, -88.98738, -88.39368, -87.95908, respectively. Then, according to formulas (9), (10), and (11), the corresponding calculations are performed. First, k1 is calculated. The calculation process is as follows:

[0079]

[0080] Then calculate x1, the calculation process is as follows:

[0081] x1=x0+k1(z1-x0)=-90+0.6(-89.15148+90)=-89.49089

[0082] Finally update the estimated error e EST(1) , the calculation formula is as follows:

[0083] e EST(1) =(1-k1)e EST(0) =(1-0.6)*3=1.2

[0084] Then, the remaining correction value, Kalman gain, and correction error are calculated in the same way. The final result is as follows: Figure 6 shown.

[0085] The current correction value when the number of operations reaches the set number is used as the pointing attitude value of the terminal turntable, and the turntable is fixed. Due to the disturbance of wind force at the working site of the equipment terminal, the data directly obtained through the main control contains low-frequency disturbance components. The present invention pre-processes the data by means of mean filtering and linear fitting, which significantly reduces the low-frequency disturbance component in the data, thereby reducing the impact of wind disturbance.

[0086] The present invention uses the Kalman filter algorithm to reduce the noise of 150 measurement values ​​obtained by step 5 to obtain the result data after pre-processing and noise reduction. At the same time, 325 groups of original data (azimuth and attitude data without any processing) are directly reduced in noise by Kalman filtering, and the filtering results of the two are compared with the original data. Figure 7 As shown, the blue curve represents direct noise reduction, and the green curve is the noise reduction result of the present invention. By comparison, it is found that the present invention performs noise reduction on 150 sets of data, and compared with directly performing noise reduction on 325 sets of data through the Kalman filter algorithm, the final noise reduction effect of both is very obvious, and the noise reduction results (stable result values) are almost the same, but the proposed solution of the present invention can achieve almost the same effect in a shorter time, and can enable the terminal turntable to point accurately in a shorter time.

Claims

1. A method for improving the initial pointing accuracy of a laser communication terminal, characterized in that: The following steps are involved: Step 1: The equipment terminal is initialized and runs, and the dual-end terminal turntable is adjusted according to the positioning data output by the GPS positioning system, so that the dual-end equipment can be initially aligned; Step 2, after step 1 is executed, the attitude data of the turntable is obtained from the device terminal through the main control, and the attitude data includes the azimuth data and pitch data of the terminal turntable; in the process of obtaining the data, the data is subjected to mean filtering operation to suppress the high-frequency interference components in the data; Step 3, according to the data result obtained by the operation in step 2, the fitting value of the group of data is obtained by using the linear fitting method; Step 4, according to the fitting value obtained in step 3, the component of low-frequency interference in the group of data is calculated by combining the fitting value with the data result obtained by the operation in step 2; Step 5, according to the posture data obtained in step 2 and the low-frequency interference component in the data obtained in step 4, calculate the posture data after removing the low-frequency interference component, and use the data as the posture measurement value after removing the low-frequency component; Step 6, based on the processed measurement value obtained in step 5, perform Kalman filter operation on it. When the number of operations reaches the set number, take the correction value obtained by the current operation as the pointing attitude value of the turntable, and fix the terminal turntable through the main control according to the obtained turntable attitude value.

2. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 1, characterized in that: The mean filtering operation in step 2 is specifically as follows: 1) Take 150 azimuth or elevation raw data, the sliding window size is 1*10, and the step size is 1. In this way, the amount of data of the calculation result is only 141. In order to keep the amount of data of the calculation result unchanged, 9 data need to be added; 2) Take 150 azimuth data, each group of data is: θ fw(1) ~θ fw(10) ,θ fw(2) ~θ fw(11) ,θ fw(3) ~θ fw(12) Until θ fw(150) ~θ fw(159) ; For each set of posture data (θ fw(j) ,θ fy(j) ) are respectively subjected to mean filtering, where θ fw is the position data of the turntable, θ fy is the pitch data of the turntable, and the mean value of each group is: (θ fwc(i) ,θ fyc(i) ), the calculation formula is: Where i is the group for mean filtering operation, i∈(0,149), j is the serial number of the supplemented data, j∈(1,159), θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

3. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 2, characterized in that: The method of supplementing data is: every 15 original data are supplemented with the mean of these 15 data, and the last 15 data do not need to be supplemented, so that 9 data will be added to the 150 original data, and the amount of data after supplementation becomes 159. At this time, the above-mentioned mean filtering operation is performed on it, and the resulting data amount becomes 150, which is the same as the original data amount before supplementation.

4. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 1, characterized in that: The fitting value (θ fwc ,θ fyc ) is calculated as: where θ fwc is the fitting value of the data after the mean filtering of the azimuth data, θ fyc is the fitting value of the pitch data after mean filtering, θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

5. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 1, characterized in that: The low-frequency interference component in the data in step 4: (θ fwcz(i) ,θ fycz(i) ), which is calculated as follows: i fwcz(i) =θ fwc(i) -θ fwc formula(5) i fycz(i) =θ fyc(i) -θ fyc formula(6) where θ fwcz(i) is the low-frequency interference component in the azimuth data, θ fycz(i) is the low-frequency interference component in the pitch data, θ fwc is the fitting value of the data after the mean filtering of the azimuth data, θ fyc is the fitting value of the pitch data after mean filtering, θ fwc(i) is the mean filtering result of the azimuth data, θ fyc(i) is the mean filtering result of the pitch data.

6. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 1, characterized in that: The final posture measurement value calculation method is: according to the original data (θ fwor(t) ,θ fyor(t) ), t is the serial number of the original posture data obtained by the master control, t∈(1,150); the low-frequency interference component (θ fwcz(t-1) ,θ fycz(t-1) ), and obtain the processed measurement value (θ fwme(t) ,θ fyme(t) ), the calculation formula is as follows: i fwme(t) =θ fwor(t)- i fwcz(t-1) formula(7) i fyme(t) =θ fyor(t)- i fycz(t-1) formula(8) where θ fwme(t) is the measured value of the orientation after processing, θ fyme(t) is the measured value of pitch after processing; θ fw o r(t) is the original orientation data, θ fyor(t) is the original data of pitch; θ fwcz(t-1) is the low-frequency interference component in the azimuth data, θ fycz(t-1) It is the low-frequency interference component in the pitch data.

7. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 1, characterized in that: The formula for the Kalman filter operation is: x k =x (k-1) +k k (z k -x (k-1) ) Formula (10) e EST(k) =(1-kk)e EST(k-1) Formula (11) e EST(k-1) is the last correction error, x (k-1) is the last correction value, z k is the current measured value, k k is the current Kalman gain constant, x (k) is the current correction value, e EST(k) is the current correction error.

8. A method for improving the initial pointing accuracy of a laser communication terminal as claimed in claim 7, characterized in that: Before the Kalman filter operation, an initial correction error e must be given according to the actual performance of the device. EST(0) , initial correction value x0 and measurement error e MEA(k) , the calculation process is divided into three steps: 1) Calculate the current Kalman gain k k ; 2) Calculate the current correction value x k ; 3) Finally calculate the current correction error e EST(k) ; A total of 150 posture data are collected through step 5, and all the data are sequentially substituted into formulas (9) to (11) to update the gain k k and the corrected error e EST(k) When the number of operations reaches the set number, the current correction value is taken as the final attitude value of the terminal turntable, and the terminal turntable is fixed according to the obtained attitude value through the main control.

Images