Partial ambiguity resolving method suitable for complex dynamic environment

Abstract

A partial ambiguity resolving method suitable for a complex dynamic environment belongs to the field of high-precision dynamic satellite navigation relative positioning, and comprises the following steps: constructing a single-epoch carrier phase double-difference difference observation equation; and calculating a weight factor corresponding to each satellite through a standardized residual error and an IGG3 method, and substituting the weight factors into the double-difference factor matrix to obtain a corrected observation value. And resolving the updated observation equation through a weighted iterative least square method. And carrying out partial ambiguity fixation on the corrected ambiguity floating point solution to obtain a whole cycle ambiguity fixed solution. And carrying out Ratio-based posteriori consistency traversal detection, monitoring the number of satellites participating in calculation in a traversal search process, and if the number of the satellites is smaller than a threshold value, outputting a floating point solution. And if the number of satellites after traversal search is greater than or equal to And substituting the fixed solution of the integer ambiguity into the updated observation equation to obtain a fixed solution of a three-dimensional relative position vector and a covariance matrix of the three-dimensional relative position vector.

Description

Technical Field

[0001] The present invention belongs to the technical field of high-precision dynamic satellite navigation relative positioning, and relates to a carrier phase differential partial ambiguity resolution method applicable to complex dynamic environments. Background Art

[0002] In view of the navigation problems during the autonomous landing process of rocket recovery and the problems of high-precision measurement of the trajectory and speed of test flight vehicles in the range test field of the aerospace measurement and control field, currently, the combined navigation method of multiple navigation devices is mainly adopted at home and abroad. Whether it is official agencies or private commercial rocket recovery, the combined navigation method of satellite navigation and other navigation modes such as inertial navigation is concentratedly used. Satellite navigation generally adopts high-precision carrier differential positioning to improve the navigation positioning accuracy. Among them, satellite navigation occupies a dominant position with its advantages of global range, applicability to different weathers, and high-precision positioning effects. Therefore, the research on carrier phase differential positioning has great practical significance in the landing navigation stage.

[0003] With the continuous increase in the demand for high-precision navigation positioning, real-time kinematic positioning has gradually become a research hotspot in recent years. The static positioning accuracy of RTK positioning can reach the millimeter level, but the success rate of ambiguity positioning has always been a major problem in RTK positioning. In the ambiguity positioning solution of RTK positioning, partial ambiguity resolution (PAR) has been widely used.

[0004] In the case of rocket recovery, especially in the high-dynamic complex environment of sea rocket recovery, the observed values of each satellite received by the receiver are not as ideal as the static data in an open environment. There are certain gross errors in the observed values of different satellites, which may have a certain impact on the resolution of ambiguities. Summary of the Invention

[0005] Aiming at the difficulty of fixing ambiguities and the influence of gross errors in observed values in complex dynamic environments, the purpose of the present invention is to provide a partial ambiguity resolution method applicable to complex dynamic environments. The improved partial ambiguity algorithm passes through triple constraints of ambiguity fixing success rate, FFRT detection, and Ratio detection, and obtains a subset of ambiguities that is easier to fix under constraints by descending order elimination of the conditional variance matrix. The present invention realizes partial ambiguity-fixed carrier phase relative positioning based on residual correction, can reduce the influence of gross errors in carrier phase observations, and has higher positioning accuracy and positioning ratio than real-time differential RTD and full ambiguity resolution FAR.

[0006] The purpose of the present invention is achieved through the following technical solutions:

[0007] A partial ambiguity resolution method applicable to complex dynamic environments disclosed by the present invention includes the following steps:

[0008] Step S11: Use a satellite antenna and a satellite navigation receiver to receive the satellite signals of the mobile station and the reference station respectively. Based on the observation data and ephemeris data of the two satellite signals, calculate the satellite positions and the receiver observation line-of-sight vectors, and construct a single-epoch carrier-phase double-difference observation equation.

[0009] Step S12: Perform a priori residual correction on the observations in S11: Calculate the weight factor corresponding to each satellite through the standardized residual and the IGG3 method, and substitute the weight factor into the double-difference factor matrix to obtain the corrected observations. Update the observation equation.

[0010] Step S13: Solve the updated observation equation in S12 by weighted iterative least squares to obtain the corrected floating-point solution and the variance matrix.

[0011] Step S14: Fix part of the ambiguities of the corrected ambiguity floating-point solution in Step S13 to obtain the integer ambiguity fixed solution.

[0012] Step S15: Perform a posteriori consistency traversal detection based on Ratio. During the traversal search process, monitor the number of satellites participating in the solution. If the number of satellites is less than the threshold, output the floating-point solution. If the number of satellites is greater than or equal to the threshold after the traversal search, proceed to the next step.

[0013] Step S16: Substitute the integer ambiguity fixed solution into the updated observation equation to obtain the fixed solution of the three-dimensional relative position vector and its covariance matrix, that is, realize the partial ambiguity solution of carrier-phase differential.

[0014] It also includes Step S17: According to the partial ambiguity solution result of carrier-phase differential obtained in Step S16, and through descending order elimination using the conditional variance matrix, obtain a subset of ambiguities that are easier to fix under constraints, thereby significantly improving the integer ambiguity fixing efficiency and being more applicable to the navigation scenario in the autonomous landing stage of rocket recovery.

[0015] The observation equation obtained in Step S11 is jointly composed of the pseudorange observations and the carrier-phase observations within one epoch, where the weight value is set to 1:10000. The specific expression of the single-epoch carrier-phase double-difference observation equation is:

[0016]

[0017] D g is the double-difference operation matrix, is the single-difference carrier-phase observation value, λ g represents the carrier wavelength of constellation g, b is the relative position vector, is the single-difference integer ambiguity, is the tropospheric error of the double-difference carrier-phase observation, is the measurement noise of the double-difference code observation value, is the current layer error.

[0018] The specific implementation method of step S13 is as follows:

[0019] Simplify the updated observation equation into a matrix expression for least squares solution:

[0020] y = A(Δb ur ) + BN + ε

[0021] where y is the double-difference carrier phase and pseudo-range measurement vector given by the receiver, A and B are constant coefficient matrices, and N is the double-difference full-cycle ambiguity vector. Δb ur is the unknown baseline vector or baseline vector correction amount.

[0022] Based on the least squares principle, the least squares solution is obtained:

[0023]

[0024] Q x = ([A B] T C[A B]) -1

[0025] where C = Q -1 , and Q is defined as the weight matrix. The weight matrix Q is determined using the measurement noise. The original observation value of the satellite carrier phase is an independent and identically distributed random variable, so the variance of the carrier phase observation value is: The variance of the pseudo-range observation value is:

[0026]

[0027] where f 1n = f1 + fn, f is the carrier phase weight, and p is the pseudo-range weight. A joint weight determination model considering the elevation angle and carrier-to-noise ratio is adopted, taking into account both the carrier-to-noise ratio factor and the satellite elevation angle factor. The specific weight determination model is as follows:

[0028]

[0029] where E i is the elevation angle of the i-th epoch satellite, C / N 0,Measured is the measured carrier-to-noise ratio.

[0030] The said step S14 includes the following steps:

[0031] Step S142: Sort the ambiguity float solution subsets according to the conditional variance matrix, and exclude the satellites with poor signal quality through the elevation angle and signal-to-noise ratio thresholds. The minimum number of satellites participating in the solution is n = 8, and the ambiguity fixing success rate threshold is P = 0.995.

[0032] Step S143: Calculate the lower limit of the ambiguity fixing success rate of the ambiguity float solution subsets by the sequential rounding method, and compare it with the preset threshold. If the success rate is greater than the threshold, go to the next step; otherwise, remove the last element in the ambiguity float solution subsets. If the number of satellites does not meet the requirements after removing the element, output the float solution; otherwise, repeat this step.

[0033] Step S144: Perform FFRT look-up table pass through the number of satellites and the ambiguity fixing success rate obtained in the previous step, and compare it with the preset threshold for FFRT detection. If FFRT is greater than the threshold, go to the next step; otherwise, remove the last element in the ambiguity float solution subsets. If the number of satellites does not meet the requirements after removing the element, output the float solution; otherwise, repeat this step.

[0034] Step S145: Search for the fixed solution using the MLAMBDA algorithm for the obtained improved partial ambiguity float solution subsets.

[0035] Step S146: Perform Ratio decision on the optimal solution and the sub-optimal solution obtained by the MLAMBDA algorithm to fix part of the ambiguities.

[0036] The said Step S15 includes the following steps:

[0037] Step S151: Adopt the elevation angle signal-to-noise ratio comprehensive weighting model for sorting to obtain the optimal 4 groups of double-difference satellites as the basic subset for posterior consistency detection, and calculate the basic Ratio value of the basic subset.

[0038] Step S152: After obtaining the basic subset, sequentially add the remaining ambiguities in the partial ambiguity subsets to the basic subset. If the Ratio value increases compared with the basic Ratio value, add this ambiguity to the basic subset to obtain a new posterior subset.

[0039] Step S153: If the dimension of the posterior subset obtained after the first traversal search < 6, output the float solution; if the posterior subset ≥ 6, perform the second traversal search. In the second traversal search, sequentially remove each ambiguity in the posterior subset to form a new posterior subset. If the Ratio value of the new posterior subset is greater than the Ratio value of the previous posterior subset, remove this satellite until the dimension of the posterior subset = 6 or the traversal search is completed.

[0040] Step S154: Output the ambiguity subset after posterior Ratio consistency and perform Mlambda algorithm fixing.

[0041] In step S13, the residual correction factor correction method is the IGG3 algorithm. Through the floating-point solution and its variance-covariance matrix, the residuals of the carrier phase double difference and the pseudorange double difference of each satellite are calculated respectively. Then, by the Danish method, the residual correction factors of each satellite are calculated using the standardized residuals, and the residual correction factors are superimposed on the weighted matrix of the least squares to recalculate the corrected floating-point solution and the variance matrix.

[0042]

[0043] where Sv i is the standardized residual of the pseudorange and the carrier phase; k 0 and k 1 are two thresholds of the standardized residual, k 0 takes the value of 1, k 1 takes the value of 3, and λ i is the weight factor of the IGG3 method weight function.

[0044] The Ratio value detection is ratio detection, which is defined as the ratio of the quadratic form of the residual of the sub-optimal solution to the optimal solution in the ambiguity fixation. Essentially, it is a discriminant detection of the distinguishability between the optimal solution and the sub-optimal solution in the ambiguity fixation. When the optimal solution is infinitely close to the true value, the distinguishability between the sub-optimal solution and the optimal solution is greater. Therefore, the accuracy of the optimal solution can be estimated by the Ratio value determination. Usually, the threshold of the Ratio detection is set manually. When the obtained Ratio value is greater than the threshold, it is considered that the ambiguity fixation is successful, and the fixed value is used as the final result. Otherwise, it is considered that the accuracy of the ambiguity fixation is poor, and the floating-point solution is used for the final result. The Ratio detection method is data-driven, and the threshold is usually set to 2-3. The specific discrimination conditions are as follows:

[0045]

[0046] where, is the floating-point solution, is the optimal solution, is the sub-optimal solution, is the covariance matrix of the floating-point solution.

[0047] Beneficial effects:

[0048] 1. A partial ambiguity resolution method applicable to complex dynamic environments disclosed by the present invention. Aiming at the complex dynamic environment of the recovery rocket's autonomous landing, a relative positioning of carrier phase with partial ambiguity fixed is realized based on residual correction. The differential positioning observation equation of the present invention is solved in a single epoch, which has better dynamic adaptability. The partial ambiguity resolution method of the present invention has better positioning accuracy and stronger robustness than the traditional full ambiguity resolution method in a poor signal environment. Compared with the traditional carrier phase differential positioning algorithm, the present invention has wider applicability and higher positioning accuracy in the complex dynamic environment of autonomous landing, and has better versatility.

[0049] 2. A partial ambiguity resolution method applicable to complex dynamic environments disclosed by the present invention. During the landing stage of the recovery rocket, a single-epoch carrier phase double-difference observation equation is constructed for long-baseline differential positioning. The Mlambda algorithm is used to fix the carrier phase differential ambiguity. For the complex dynamic application scenario of the recovery rocket landing on the ship, through the success rate of ambiguity fixing, FFRT detection, and Ratio detection of the comprehensive pre-observation residual correction and post-observation consistency detection, based on the triple-constraint partial ambiguity resolution method, the ambiguity subset that is easier to fix under constraints is obtained by descending order elimination of the conditional variance matrix, reducing the influence of gross errors in carrier phase observations. Compared with the traditional LAMBDA algorithm, the probability of incorrect fixing is significantly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flowchart of a partial ambiguity resolution method applicable to complex dynamic environments of the present invention.

[0051] Figure 2 A flowchart of the MLAMBDA algorithm of the present invention.

[0052] Figure 3 The geometric relationship between the satellite and the receiver. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The following examples are given in conjunction with the drawings to describe the present invention in detail.

[0054] As Figure 1 shown, a partial ambiguity resolution method applicable to complex dynamic environments disclosed in this embodiment is specifically implemented as follows:

[0055] Step S11: Use the satellite antenna and the satellite navigation receiver to receive the satellite signals of the mobile station and the reference station respectively, and perform acquisition and tracking processing. Calculate the satellite position and the receiver's observation direction vector through the satellite signal observation data and ephemeris data obtained by the satellite navigation receiver, and construct a carrier phase double-difference model.

[0056]

[0057]

[0058] Where: λ s represents the carrier wavelength of constellation g; represents the fractional part of the single-difference carrier phase observation value of the s-th satellite in constellation g; represents the integer cycle part of the single-difference carrier phase observation value of the s-th satellite in constellation g; represents the single-difference code observation value of the s-th satellite in constellation g; represents the measurement noise of the single-difference carrier phase observation value of the s-th satellite in constellation g; represents the measurement noise of the single-difference code observation value of the s-th satellite in constellation g; represents the single-difference value of the station-satellite distance of antenna u and antenna r with respect to the s-th satellite in constellation g; dt u represents the clock error of the receiver to which antenna u belongs; dt r represents the clock error of the receiver to which antenna r belongs; c represents the speed of light.

[0059] Step S12, solve the observation equation by weighted iterative least squares method: The observation equation of single-epoch carrier phase double difference can be expressed as:

[0060]

[0061] D g is the double-difference operation matrix, is the single-difference carrier phase observation value, λ g represents the carrier wavelength of constellation g, b is the relative position vector, is the single-difference integer ambiguity, is the tropospheric error of the double-difference carrier phase observation quantity, is the measurement noise of the double-difference code observation value, is the ionospheric error.

[0062] For easy understanding, the double-difference observation equation can be simplified to the matrix expression for least squares solution:

[0063] y = A(Δb ur ) + BN + ε

[0064] where y is the double-difference carrier phase and pseudo-range measurement vector given by the receiver, A and B are constant coefficient matrices, N is the double-difference full-cycle ambiguity vector, and Δb ur is the unknown baseline vector or baseline vector correction quantity. Based on the least squares principle, the least squares solution can be obtained:

[0065]

[0066] Q x = ([A B] T C[A B]) -1

[0067] where C = Q -1 , and Q is defined as the weight matrix. The weight matrix Q is determined using measurement noise. The original observations of the satellite carrier phase are independent and identically distributed random variables, and the variance of the carrier phase observations is: The variance of the pseudorange observations is:

[0068]

[0069] where f 1n = f1 + fn, f is the carrier phase weight, and p is the pseudorange weight. A combined weight determination model based on elevation angle and carrier-to-noise ratio is adopted, taking into account both the carrier-to-noise ratio factor and the satellite elevation angle factor. The specific weight determination model is as follows:

[0070]

[0071] where E i is the elevation angle of the satellite at the i-th epoch, and C / N 0,Measured is the measured carrier-to-noise ratio.

[0072] In this step, the floating-point solutions of the relative positioning vector and the double-difference ambiguity are obtained.

[0073] Step S13: Perform prior residual correction on the satellite signal observations. Calculate the weight factor corresponding to each satellite through the standardized residual and the IGG3 method, and substitute the weight factor into the double-difference factor matrix. Finally, obtain the corrected observations, perform the solution, and obtain the corrected floating-point solution and the variance matrix.

[0074] In the present invention, the ambiguity correction model based on prior residuals is, in the least squares iteration, through the floating-point solution and its variance-covariance matrix, calculate the residuals of the carrier phase double-difference and the pseudorange double-difference of each satellite respectively, and then through the Danish method, calculate the residual correction factor of each satellite using the standardized residual, and superimpose the residual correction factor on the weighted matrix of the least squares to recalculate the corrected floating-point solution and the variance matrix. The IGG3 method is a measurement robust estimation theory, and the model is as follows:

[0075]

[0076] where Sv i is the standardized residual of the pseudorange and the carrier phase; k 0 and k 1 are two thresholds of the standardized residual. In this article, k 0 takes the value of 1, and k 1The value is 3, λ i is the weight factor of the IGG3 legal right function.

[0077] Step S14: Fix part of the ambiguities in the corrected ambiguity subset in step S13;

[0078] In the present invention, only when the integer value of the ambiguity is fixed can a high-precision and highly reliable baseline vector be obtained. When the integer value of the ambiguity is fixed incorrectly, the result of the baseline vector will have a greater error. The fixing of the integer ambiguity uses the least squares ambiguity decorrelation adjustment method to search and fix.

[0079] As Figure 2 shown, in some embodiments, step S14 includes: step S141: Convert the obtained ambiguity subset to LDL decomposition in the Z space, perform ambiguity decorrelation, and obtain the conditional variance matrix. There are mainly two tasks in the decorrelation stage:

[0080] (1) Make the absolute value of the non-diagonal elements in the L matrix less than 0.5.

[0081] (2) Sort the diagonal elements in the D matrix in descending order.

[0082] In some embodiments, step S14 includes: step S142: Sort the ambiguity subset according to the conditional variance matrix, and exclude satellites with poor signal quality through the elevation angle and signal-to-noise ratio thresholds.

[0083] In some embodiments, step S14 includes: step S143: Calculate the lower limit of the ambiguity fixing success rate of the ambiguity subset by the sequential rounding method, and compare it with the set threshold. If the success rate is greater than the threshold, proceed to the next step; otherwise, remove the last element in the ambiguity subset. If the number of satellites does not meet the requirements after removing the element, output the float solution; otherwise, repeat this step.

[0084] In some embodiments, step S14 includes: step S144: Perform FFRT detection by looking up the FFRT table and comparing it with the set threshold. If the FFRT is greater than the threshold, proceed to the next step; otherwise, remove the last element in the ambiguity subset. If the number of satellites does not meet the requirements after removing the element, output the float solution; otherwise, repeat this step.

[0085] In some embodiments, step S14 includes: step S145: For the obtained improved partial ambiguity subset, perform MLAMBDA algorithm search for the fixed solution. The search condition for the integer ambiguity based on integer least squares can be expressed as:

[0086]

[0087] The objective function can be simplified to:

[0088]

[0089] Let the maximum value of the objective function be r 2

[0090]

[0091] To fix the ambiguity, that is, to find a set of integer ambiguity parameters that minimize the objective function f within an n-dimensional sphere with a radius of r The shape of the search space is an ellipsoid in a multi-dimensional space. The larger the radius, the larger the search space, and thus the greater the confidence of the optimal integer ambiguity obtained, but the greater the computational complexity; on the contrary, when the radius is smaller, the search space is smaller, and it is possible that the optimal integer ambiguity is not within the searched area, resulting in a smaller confidence. Since is an upper triangular matrix, the sequential solution method can be used to solve it backward starting from the last integer ambiguity, and finally solve According to the objective function, it can be obtained that

[0092]

[0093] The solution is All integers within the interval form the alternative group of integer ambiguities Bring the alternative group into the following formula, and the alternative group of the previous ambiguity can be obtained: The general formula for iteration is:

[0094]

[0095] If the range beyond the search radius appears during the calculation, the current result will be excluded. After finally obtaining the fixed group and alternative group of ambiguity parameters, bring them into the objective function. At this time, the combination that makes the objective function obtain the minimum value is the best ambiguity combination.

[0096] In some embodiments, step S14 includes: step S146, performing Ratio decision on the optimal solution and the sub-optimal solution obtained by the MLAMBDA algorithm.

[0097] Step S15, performing posterior consistency traversal detection based on Ratio. During the traversal search process, monitor the number of satellites participating in the solution. If the number of satellites is less than the threshold, output the floating-point solution. If the number of satellites is greater than or equal to the threshold after the traversal search, proceed to the next step;

[0098] In the present invention, the posterior consistency detection based on Ratio is for the extreme case where the residuals of some satellite observations are relatively large after the satellite loses lock. By using the ambiguity subset and covariance matrix obtained by fixing some ambiguities based on the correction of the prior residuals, a set of ambiguity-fixed solutions with the optimal Ratio value is obtained through traversal search.

[0099] In some embodiments, step S15 includes: Step S151, adopting a comprehensive weight determination model of elevation angle signal-to-noise ratio for sorting to obtain the optimal 4 groups of double-difference satellites as the basic subset for posterior consistency detection, and calculating the basic Ratio value of the basic subset.

[0100] In some embodiments, step S15 includes: Step S152, after obtaining the basic subset, sequentially add the remaining ambiguities in the partial ambiguity subset to the basic subset. If the Ratio value increases compared to the basic Ratio value, add this ambiguity to the basic subset to obtain a new posterior subset.

[0101] In some embodiments, step S15 includes: Step S153, if the dimension of the posterior subset obtained after the first traversal search < 6, output the floating-point solution. If the posterior subset ≥ 6, perform a second traversal search. In the second traversal search, sequentially remove each ambiguity in the posterior subset to form a new posterior subset. If the Ratio value of the new posterior subset is greater than the Ratio value of the previous posterior subset, remove this satellite until the dimension of the posterior subset = 6 or the traversal search is completed.

[0102] In some embodiments, step S15 includes: Step S154, output the ambiguity subset after posterior Ratio consistency and perform the Mlambda algorithm for fixing.

[0103] Step S16, obtain the fixed solution of the integer ambiguity, the fixed solution of the three-dimensional relative position vector, and its covariance matrix.

[0104]

[0105] It further includes step S17, according to the carrier phase differential partial ambiguity solution result obtained in step S16, and eliminating by descending order of the conditional variance matrix to obtain an ambiguity subset that is easier to fix under constraints, thereby significantly improving the integer ambiguity fixing efficiency and being more applicable to the navigation scenario in the autonomous landing stage of rocket recovery.

[0106] The above specific description further details the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above is only the specific embodiment of the present invention and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A partial ambiguity resolution method suitable for complex dynamic environments, characterized by: The following steps are involved: Step S11, using a satellite antenna and a satellite navigation receiver to receive satellite signals from a mobile station and a base station respectively, calculating the satellite position and the receiver observation line-of-sight vector through the observation data and ephemeris data of the two satellite signals, and constructing a single epoch carrier phase double-difference differential observation equation; Step S12, perform a priori residual correction on the observation value of S11: calculate the corresponding weight factor of each satellite by standardized residual and IGG3 method, and bring the weight factor into the double difference factor matrix to obtain the corrected observation value; update the observation equation; Step S13, solving the observation equation updated in S12 by weighted iterative least squares method, and obtaining a corrected floating point solution and variance matrix by using the least squares method; Step S14, fixing the partial ambiguity of the ambiguity floating point solution corrected in step S13 to obtain an integer ambiguity fixed solution; Step S15: Perform a posteriori consistency traversal test based on Ratio. During the traversal search, monitor the number of satellites involved in the solution. If the number of satellites is less than the threshold, output a floating point solution. If the number of satellites is greater than or equal to the threshold after the traversal search, proceed to the next step. Step S16, bring the fixed solution of the integer ambiguity into the updated observation equation to obtain the fixed solution of the three-dimensional relative position vector and its covariance matrix, that is, realize the carrier phase difference partial ambiguity resolution.

2. A partial ambiguity resolution method suitable for complex dynamic environments as claimed in claim 1, characterized in that: The observation equation obtained in step S11 is composed of the pseudorange observation value and the carrier phase observation value within one epoch, wherein the weight is set to 1:10000; the single epoch carrier phase double difference observation equation is specifically expressed as: D g is the double difference operation matrix, is the single-difference carrier phase observation, λ g represents the carrier wavelength of constellation g, b is the relative position vector, is the single-difference integer ambiguity, is the tropospheric error of the double-difference carrier phase observation, is the measurement noise of the double-differenced observations, is the current layer error.

3. A partial ambiguity resolution method suitable for complex dynamic environments as claimed in claim 2, characterized in that: The specific implementation method of step S13 is: The updated observation equation is simplified to a matrix expression solved by the least squares method: y=A(Δb ur )+BN+ε Where y is the double difference carrier phase and pseudo-range measurement vector given by the receiver, A and B are constant coefficient matrices, and N is the double difference full-cycle ambiguity vector; Δb ur is the unknown baseline vector or baseline vector correction; Based on the least squares principle, the least squares solution is obtained: Q x =([A B] T C[A B]) -1 Where C = Q -1 , Q is defined as the weight matrix; the weight matrix Q is determined using measurement noise, and the original observation value of the satellite carrier phase is an independent equidistributed random variable, so the variance of the carrier phase observation value is: The variance of the pseudorange observations is: where f 1n =f1+fn, f is the carrier phase weight, p is the pseudorange weight; the joint weighting model of altitude angle and carrier-to-noise ratio is adopted, which takes into account both the carrier-to-noise ratio factor and the satellite altitude angle factor. The specific weighting model is as follows: Where E i is the altitude angle of the satellite at the ith epoch, C / N 0,Measured is the measured carrier-to-noise ratio.

4. A partial ambiguity resolution method suitable for complex dynamic environments as claimed in claim 3, characterized in that: The step S14 comprises the following steps: Step S142, sorting the ambiguity floating point solution subsets according to the conditional variance matrix, and excluding satellites with poor signal quality by using the altitude angle and signal-to-noise ratio thresholds, the minimum number of satellites involved in the solution n = 8, and the ambiguity fixation success rate threshold P = 0.995; Step S143, calculate the lower limit of the ambiguity fixing success rate of the ambiguity floating point solution subset by the sequential rounding method, and compare it with the preset threshold. If the success rate is greater than the threshold, proceed to the next step, otherwise, remove the last element in the ambiguity floating point solution subset. After removing the element, if the number of satellites does not meet the requirement, output the floating point solution, otherwise repeat this step; Step S144, perform FFRT table lookup based on the number of satellites and the ambiguity fixation success rate obtained in the previous step, and compare with the preset threshold for FFRT detection. If the FFRT is greater than the threshold, proceed to the next step, otherwise, remove the last element in the ambiguity floating-point solution subset. After removing the element, if the number of satellites does not meet the requirement, output the floating-point solution, otherwise repeat this step; Step S145, the obtained improved partial ambiguity floating point solution subset is subjected to MLAMBDA algorithm to search for a fixed solution; Step S146, performing Ratio judgment on the optimal solution and the suboptimal solution obtained by the MLAMBDA algorithm to achieve partial ambiguity fixation.

5. A partial ambiguity resolution method suitable for complex dynamic environments as claimed in claim 4, characterized in that: The step S15 comprises the following steps: Step S151, using the altitude angle signal-to-noise ratio comprehensive weighting model to sort, obtain the best 4 groups of double-difference satellites as the basic subset for a posteriori consistency detection; and calculate the basic Ratio value of the basic subset; Step S152: after obtaining the basic subset, the remaining ambiguities in the partial ambiguity subset are sequentially added to the basic subset. If the Ratio value is larger than the basic Ratio value, the ambiguity is added to the basic subset to obtain a new posterior subset. Step S153: if the dimension of the a posteriori subset obtained after the first traversal search is less than 6, a floating point solution is output; if the a posteriori subset is ≥ 6, a second traversal search is performed, and each ambiguity in the a posteriori subset is removed in turn in the second traversal search to form a new a posteriori subset; if the Ratio value of the new a posteriori subset is greater than the Ratio value of the previous a posteriori subset, the satellite is removed until the dimension of the a posteriori subset is 6 or the traversal search is completed; Step S154: output the fuzzy subset after the posterior Ratio consistency, and fix it using the Mlambda algorithm.

6. A partial ambiguity resolution method suitable for complex dynamic environments as claimed in claim 5, characterized in that: In the step S13, the residual correction factor correction method used is the IGG3 algorithm; the residuals of the carrier phase double difference and the pseudorange double difference of each satellite are calculated respectively through the floating point solution and its variance covariance matrix, and then the residual correction factor of each satellite is calculated by the Danish method using the standardized residual, and the residual correction factor is superimposed on the weighted matrix of the least squares, and the corrected floating point solution and variance matrix are recalculated; Among them Sv i is the standardized residual of pseudorange and carrier phase; k0 and k1 are two thresholds of standardized residual. In this paper, k0 is 1, k1 is 3, and λ i is the weight factor of the IGG3 weight function.

7. A partial ambiguity resolution method suitable for complex dynamic environments as claimed in claim 6, characterized in that: The Ratio value detection is a ratio detection, which is defined as the ratio of the residual quadratic form of the suboptimal solution and the optimal solution in the ambiguity fixation. When the optimal solution is infinitely close to the true value, the distinction between the suboptimal solution and the optimal solution is greater, so the accuracy of the optimal solution can be estimated by the Ratio value judgment. Usually, the threshold of the Ratio detection is manually set. When the obtained Ratio value is greater than the threshold, it is considered that the ambiguity is fixed successfully, and the fixed value is used as the final result. Otherwise, it is considered that the ambiguity fixation accuracy is poor, and the floating point solution is used for the final result. The Ratio detection method is data-driven, and the threshold is usually set to 2-3. The specific judgment conditions are as follows: in, For floating point solutions, is the optimal solution, is a suboptimal solution, Solved covariance matrix for floating point.

8. A partial ambiguity resolution method suitable for use in a complex dynamic environment as claimed in claim 1, 2, 3, 4, 5, 6 or 7, characterized in that: It also includes step S17, according to the carrier phase difference partial ambiguity resolution result obtained in step S16, and eliminates it in descending order through the conditional variance matrix to obtain an ambiguity subset that is easier to fix under the constraint, thereby significantly improving the efficiency of integer ambiguity fixation, and is more suitable for the navigation scenario of the rocket recovery autonomous landing phase.

Images