A launch vehicle program angle fitting endpoint optimization method

Abstract

The present application relates to a kind of launch vehicle program angle fitting endpoint optimization method.Different from the method of traditional fitting endpoint optimization by artificial debugging, the method is by the second derivative of standard trajectory program angle with respect to time, the larger point in it is filtered out, and is arranged combination, then the fitting accuracy of all arrangement combinations is judged by self-defined fitting accuracy performance index, so as to filter out the optimal fitting endpoint value.In addition, to ensure the fitting accuracy, adjustment strategy when the fitting accuracy is insufficient is also set, for example, when the maximum fitting angle deviation exceeds the preset maximum deviation, the fitting accuracy is improved by increasing the number of fitting endpoints, and the endpoint value is fine-tuned when the numerical jump at the fitting endpoint is too large.Through the above steps, the endpoint value obtained by the method optimization can meet the fitting accuracy requirement, and manual intervention can be completely avoided, so as to reduce the human operation.

Description

Technical Field

[0001] This invention relates to the field of launch vehicle trajectory design technology, and more specifically to a launch vehicle program angle fitting endpoint optimization method. Background Technology

[0002] After the launch vehicle's trajectory design is completed, the rocket's flight attitude angles need to be output for the launch vehicle's flight control software to use during flight, so that the rocket flies according to the predetermined attitude angles. These attitude angles are also called program angles. However, due to objective limitations such as the storage capacity of the rocket's onboard computer, it is difficult to load the program angles corresponding to every moment calculated during the standard trajectory design into the onboard computer. Therefore, in actual launch vehicle flights, the program angles output from the standard trajectory are often simplified. The simplification methods mainly include polynomial fitting and linear interpolation.

[0003] Polynomial fitting approximates the changes in rocket flight program angles (mainly pitch and yaw angles) over time (or altitude) as an nth-degree polynomial (a cubic polynomial is generally sufficient for fitting accuracy). The main advantage of this method is that the program angles can be expressed analytically and require less storage space. The disadvantage is that since the program angles in standard ballistic trajectory design are not strictly polynomial curves, piecewise fitting may be necessary to ensure fitting accuracy. However, the fitting accuracy is highly dependent on the values ​​of the selected endpoints. Currently, after calculating the second derivative of the program angle, points with significant jumps are manually selected as fitting endpoints. If the difference is large, other jump points are manually replaced or fine-tuned in their vicinity.

[0004] Linear interpolation uses a table of standard ballistic outputs that varies with time (or altitude). During actual flight, the flight control software performs linear interpolation between the current flight time (or altitude) and this table to calculate the standard ballistic program angle corresponding to the current time (or altitude). The main advantage of this method is its high accuracy when the table interval is sufficiently small. However, its disadvantage is that its accuracy is highly dependent on the table interval, and it requires a large amount of storage space when the table interval is small.

[0005] To reduce the amount of storage space used by the rocket's onboard computer and improve the accuracy of the program angle, a program angle fitting endpoint optimization method based on polynomial fitting is proposed. This method can improve the accuracy of polynomial fitting program angles while minimizing human intervention and reducing the amount of storage space used by the rocket's onboard computer. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide a program angle segmentation endpoint optimization algorithm that is simple, streamlined, and has high fitting accuracy.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: a method for optimizing endpoints of launch vehicle program angle fitting, comprising:

[0008] The second difference of the program angle with respect to time in the standard trajectory is obtained by taking the second difference of the program angle with respect to time at each moment in the input time sequence.

[0009] Select Num moments with large absolute values ​​of second-order differences; combine the moments with large absolute values ​​of second-order differences, calculate the fitting accuracy for each combination, and obtain the combination with the highest fitting accuracy.

[0010] Find the maximum angular deviation of the combination with the highest fitting accuracy. If the angular deviation does not meet the requirements, increase the number of fitting endpoints, recombine and optimize.

[0011] If the combination with the highest fitting accuracy meets the angular deviation requirement, then calculate the program angular jump value at each segment fitting endpoint. If the jump value of a certain endpoint does not meet the requirement, then adjust that endpoint until the jump values ​​of all its endpoints meet the accuracy requirement, and obtain the final endpoint.

[0012] Furthermore, the step of numerically differentiating the sequence of program angles with respect to time in a standard trajectory with respect to time to obtain the second-order difference of the program angle with respect to time at each moment in the input time sequence includes:

[0013] Let the program angle sequence be... The corresponding times are [t1 t2…t n ], take the first difference of it as The second difference is divided into The time intervals corresponding to the second-order difference are [t2 t3…t] n-1 ].

[0014] Further, the step of selecting Num moments with large absolute values ​​of second-order differences; combining these moments with large absolute values ​​of second-order differences; calculating the fitting accuracy for each combination; and obtaining the combination with the highest fitting accuracy includes the following steps:

[0015] S21. Initialize the number of points Num to be selected, and set the number of points to be filtered Num_out:

[0016] S22. Calculate the absolute value of the array ddPhi to obtain the array |ddPhi|. Get the first Num values ​​with larger values ​​in |ddPhi|, denoted as the array Arr_ddPhi. And the time corresponding to the Num values, denoted as the array Arr_TddPhi.

[0017] S23. Iterate through all numbers in the array Arr_ddPhi, compare each number with the numbers within its vicinity of T seconds, and record the number with the smaller value in each comparison.

[0018] S24. After the comparison is completed, all the smaller values ​​are removed from the records to obtain all the larger values ​​in the array Arr_ddPhi that are not repeatedly selected within T seconds.

[0019] S25. If the number of elements selected by step S14 is less than Num_out, then let Num = Num + K1, and return to step S22 to re-select until the number of elements selected by step S24 is not less than Num_out. Record the combination of the selected elements at the corresponding time as an array Arr_TddPhi_out, and record the number of elements in it as n1.

[0020] Furthermore, K1 is an integer between 5 and 50, preferably an integer between 8 and 15.

[0021] Furthermore, the step of combining the moments with larger absolute values ​​of the second-order differences, calculating the fitting accuracy for each combination, and obtaining the combination with the highest fitting accuracy includes the following steps:

[0022] S26. Set the initial value of the number of selected time points m1;

[0023] S27. Select m1 elements from Arr_TddPhi_out and combine them to obtain a total of Combinations;

[0024] S28, Regarding Arr_TddPhi_out The process iterates through all possible combinations, calculating the fitting accuracy performance index for each combination, including the following steps:

[0025] S281. Assume the current combination is [tp1 tp2…tp m1 ], will [tp1 tp2…tp m1 ] as endpoint pairs [t1 t2…t n Divide into segments;

[0026] S282. Perform cubic polynomial fitting on each segment of the program angle Phi sequence;

[0027] S283. Obtain [t1 t2…t] using the fitted polynomial. n The fitting angle result for each point is denoted as [the result is not specified in the original text].

[0028] S284. Calculate the fitting performance index under the current combination.

[0029] S29. The fitting performance indices for all combinations are as follows: The combination that yields the minimum value is the one with the highest fitting accuracy.

[0030] Furthermore, the step of finding the maximum angular deviation corresponding to the combination with the highest fitting accuracy, if the angular deviation does not meet the accuracy requirements, involves increasing the number of fitting endpoints, recombining and optimizing, including the following steps:

[0031] S31. Among the combinations with the highest fitting accuracy, find... The maximum value in;

[0032] S32, if The maximum value exceeds the allowable range of fitting deviation. Then make m1 = m1 + K2, and repeat steps S22-S32 to start the combination optimization again.

[0033] Furthermore, K2 is an integer between 1 and 5, preferably 1.

[0034] Furthermore, the process involves calculating the program angle jump value at each segmented fitting endpoint for the highest accuracy combination that meets the angular deviation accuracy requirement. If the jump value at a certain endpoint exceeds the accuracy requirement, the endpoint is adjusted until it meets the accuracy requirement. This includes the following steps:

[0035] S41. For the endpoint values ​​of the combination with the highest fitting accuracy, traverse the endpoints in ascending order of time, solve the polynomials obtained by fitting the two segments before and after each endpoint, and obtain the absolute value of the difference between the two fitting results, which is the jump value.

[0036] S42. If the jump value of a certain endpoint in the current traversal is greater than the allowed jump value. The endpoint value is then adjusted, including the following steps:

[0037] S421. Move the endpoint value in the ±y time interval with a step size of x, while keeping other endpoint values ​​unchanged, to obtain z adjustment terms;

[0038] S422. After adjusting the endpoints, a total of z+1 combinations are obtained. The fitting performance index is solved for these z+1 combinations.

[0039] S423. Find the minimum performance index of z+1 combinations. The adjusted endpoint corresponding to the minimum performance index is the final endpoint.

[0040] Furthermore, x is 0.01-0.1, y is 0.1-1, and z is 20.

[0041] Furthermore, the allowable range of the fitting deviation The maximum jump value is 0.02-0.1. It ranges from 0.01 to 0.05.

[0042] Compared with existing technologies, the advantages of this invention are:

[0043] This invention proposes an endpoint optimization method for program angle fitting of launch vehicles. Addressing the need for high-precision expression of rocket flight program angles with minimal data volume during launch missions, the method performs second-order differencing on a standard trajectory, identifies points with large second-order differencing values, and optimizes performance indicators through permutations and combinations of these points until the program angle fitting accuracy requirements for the entire flight are met with the minimum number of endpoints (i.e., the minimum number of polynomial piecewise fitting segments). Compared to traditional methods, this invention features a simple algorithm, avoids manual intervention in calculations through a programmed operation process, and ensures computational accuracy, thus meeting the program angle fitting accuracy requirements for launch vehicles. Detailed Implementation

[0044] The present invention will be further described below with reference to the embodiments.

[0045] This invention provides a method for optimizing endpoints of launch vehicle program angle fitting, the method comprising the following steps:

[0046] S1. The second-order difference of the sequence of program angles (pitch angle, yaw angle) with respect to time in the standard trajectory is obtained by numerical method. The second-order difference of program angle with respect to time at each moment in the input time sequence is obtained, which is used to approximate the second derivative of program angle with respect to time at each moment.

[0047] S2. Sort the second-order differences at each time step and select the Num time steps with the largest absolute values ​​of the second-order differences; that is, sort them from largest to smallest and select the first Num time steps.

[0048] The Num selected moments with large absolute values ​​of second-order differences are combined, and the fitting accuracy corresponding to each combination is calculated. The fitting accuracy of all combinations is sorted to obtain the combination with the highest accuracy.

[0049] S3. Find the maximum angular deviation of the fit corresponding to the combination with the highest accuracy. If the angular deviation does not meet the requirements, increase the number of fitting endpoints, re-combine and optimize.

[0050] S4. If the combination with the highest fitting accuracy meets the angular deviation requirement, then calculate the program angular jump value at each segment fitting endpoint. If the jump value of an endpoint exceeds the accuracy requirement, then adjust the endpoint until it meets the accuracy requirement and obtain the final endpoint.

[0051] The following is a detailed description of steps S1-4:

[0052] This invention provides a simple calculation method for optimizing the program angle fitting endpoints of launch vehicles, which requires no manual intervention and can ensure fitting accuracy while minimizing the amount of binding data.

[0053] The launch vehicle program angle fitting endpoint optimization method in this invention differs from the traditional method of manually adjusting and optimizing fitting endpoints. This method involves taking the second derivative of the standard ballistic program angle with respect to time, selecting the larger points, arranging and combining them, and then evaluating the fitting accuracy of all arrangements and combinations using a self-defined fitting accuracy performance index, thereby selecting the optimal fitting endpoint value.

[0054] Furthermore, to ensure fitting accuracy, an adjustment strategy is implemented when the fitting accuracy is insufficient. For example, when the maximum fitting angle deviation exceeds the preset maximum deviation, the fitting accuracy is improved by increasing the number of fitting endpoints; when the values ​​at the fitting endpoints change too drastically, the endpoint values ​​are fine-tuned. Through these steps, the endpoint values ​​optimized by this method can be guaranteed to meet the fitting accuracy requirements, and manual intervention can be completely avoided, reducing human error.

[0055] In this embodiment, the specific steps for obtaining the second-order difference of the program angle (taking elevation angle as an example) with respect to time in a standard trajectory using a numerical method to calculate the second-order difference of the program angle with respect to time at each moment in the input time series include:

[0056] Find the second difference of the sequence of program angles over time:

[0057] Assume the pitch angle sequence is The corresponding times are [t1 t2...t n (In this embodiment, the input time interval is 0.01s), and the difference is calculated once.) The second difference is divided into Since the array length is reduced after the second difference is calculated, the corresponding time after the second difference is denoted as [t2 t3...t]. n-1 ].

[0058] In this embodiment, the specific steps for selecting the moments with larger absolute values ​​of Num second-order differences include:

[0059] S21. Initialize the number of points Num and the point selection result sortTPhi, and set the number of filtered points Num_out; the specific values ​​of each parameter in this embodiment are as follows:

[0060] Num = 50

[0061] sortTPhi=0

[0062] Num_out = 10

[0063] The sortTPhi result is set up for ease of programming implementation. It is an array obtained by filtering in step 24 below. This array is assigned to sortTPhi. Then, it is determined whether the length of the sortTPhi array is greater than or equal to Num_out to determine whether a loop is needed. The first time the loop runs, it needs an initial value, otherwise it cannot run. Therefore, the initial value of sortTPhi is set to 0.

[0064] S22. Take the absolute value of the array ddPhi, that is, iterate through each element in the array. If the number is less than 0, output the opposite of the number. Otherwise, output the number directly. Finally, combine the output results to obtain the array |ddPhi|. Sort the values ​​in |ddPhi| to get the first Num values ​​with the largest values ​​in |ddPhi|, which is denoted as the array Arr_ddPhi. The time corresponding to the Num values ​​is denoted as the array Arr_TddPhi. The length of the above two arrays is Num.

[0065] S23. Iterate through all numbers in the array Arr_ddPhi, comparing each number with its neighboring numbers one by one, and recording the number with the smaller value in each comparison; the neighboring numbers refer to the numbers whose difference between the two at corresponding times is less than T seconds; in the actual calculation process, Arr_ddPhi is traversed one by one, and for each element, it is compared with all other numbers in arr_ddphi one by one. If the difference between the two at corresponding times is less than T seconds, the element with the smaller value in this comparison is recorded. Based on experience, T = 2 in this embodiment.

[0066] S24. After the comparison is completed, remove all the elements with smaller values ​​recorded, thus obtaining all the elements in the Arr_ddPhi array with larger values ​​that are not repeatedly selected within T seconds. To facilitate programming implementation, identical items can be merged before removal.

[0067] S25. If the number of elements selected by step S24 is less than Num_out, then let Num = Num + K1, and return to S2 to re-select until the number of elements selected by step S24 is not less than Num_out. Record the time corresponding to the selected array as Arr_TddPhi_out, and record its number of elements as n1. In this embodiment, K1 = 10, but in other embodiments it can be 5, 8, 15, 20, 30, 50, etc.

[0068] The specific steps for combining the moments with larger |ddPhi| values, calculating the fitting accuracy for each combination, sorting the fitting accuracies of all combinations, and obtaining the combination with the highest fitting accuracy include:

[0069] S26. Set the initial value of the number of selected time points m1. In this embodiment, the initial value m1 = 6.

[0070] S27. Select m1 elements from Arr_TddPhi_out and combine them, totaling... Combinations;

[0071] S28, Regarding Arr_TddPhi_out We iterate through all possible combinations, and for each combination, we calculate the corresponding fitting accuracy performance index. The specific steps are as follows:

[0072] S281. Assume the current combination is [tp1 tp2…tp m1 ], will [t1 t2…t n Segment according to this combination (if a time series value falls exactly at an endpoint, it is merged into the previous segment);

[0073] S282. Perform a 3rd-order polynomial fitting on each segment of the pitch angle Phi sequence.

[0074] S283. Obtain [t1 t2...t] using the fitted polynomial. n The fitted pitch angle result for each point is denoted as […].

[0075] S284. Calculate the fitting performance index under the current combination.

[0076] S29. The performance indicators for all combinations are as follows: The combination of the minimum values ​​is the one with the highest fitting accuracy.

[0077] The specific steps for finding the maximum angular deviation of the combination with the highest fitting accuracy, and then increasing the number of fitting endpoints to re-combine and optimize, are as follows:

[0078] S31. Find the combination with the highest fitting accuracy corresponding to the maximum fitting error, that is, find the combination with the smallest performance index. The maximum value in the range is used to obtain the maximum fitting angular deviation;

[0079] S32. If the maximum value exceeds the allowable range of the fitting angle deviation. ( Preferably, it is between 0.02 and 0.1. In this embodiment... If the value of m1 is increased so that m1 = m1 + K2, the combination and optimization will start again until the maximum fitting angle deviation meets the accuracy requirements. In this embodiment, K2 = 1, but in other embodiments it can be 2, 3, 4 or 5.

[0080] The specific steps for determining the program angle jump value at each segmented fitting endpoint of the combination that meets the angular deviation accuracy requirement are as follows: If the jump value at a certain endpoint exceeds the accuracy requirement, the endpoint is adjusted until the accuracy requirement is met.

[0081] S41. For the endpoint values ​​of the combination with the highest fitting accuracy, traverse the endpoints in ascending order of time, solve the polynomials obtained by fitting the two segments before and after each endpoint, and obtain the absolute value of the difference between the two fitting results.

[0082] S42. If the difference between the fitting results of two segments before and after a certain endpoint of the current traversal is greater than the maximum jump value. ( Preferably, it is between 0.01 and 0.05. In this embodiment... If the endpoint value is not specified, then the value of that endpoint will be adjusted. The specific steps are as follows:

[0083] S421. Move the endpoint value within the range of ±y with a step size of x to obtain z adjustment items; the value of x is preferably between 0.01 and 0.1, and the value of y is preferably between 0.1 and 1; in this embodiment, x is 0.01 and y is 0.1 (i.e., add adjustment items of [-0.1 -0.09 -0.08...0.08 0.09 0.1] to the endpoint value), resulting in 20 adjustment items. The adjustment range and step size can be determined according to specific needs. In other embodiments, the value of x can be 0.05 and the value of y can be 0.6; during adjustment, except for the endpoint that needs to be adjusted, the other endpoint values ​​remain unchanged;

[0084] S422. After the endpoints are adjusted, there are a total of 21 combinations (20 adjustable endpoints plus 1 unadjusted original endpoint, each endpoint corresponds to a combination). The fitting performance index of these 21 combinations is solved using the same method as above.

[0085] S423. Optimize the performance index of the 21 combinations, find the minimum value, and the adjusted endpoint corresponding to the minimum performance index is the value of the adjusted endpoint, thus obtaining the final endpoint.

[0086] In summary, this invention proposes an endpoint optimization method for launch vehicle program angle fitting. Addressing the need for expressing the rocket's flight program angle with the highest possible accuracy using minimal data in launch missions, this method performs second-order differencing on a standard trajectory, selects points with large second-order differencing values, and optimizes these points through permutations and combinations to find the optimal performance indicators until the required program angle fitting accuracy for the entire flight is met with the minimum number of endpoints (i.e., the minimum number of polynomial piecewise fitting segments). Compared to traditional methods, this invention features a simple algorithm, avoids manual intervention in calculations through a standardized operational procedure, and ensures computational accuracy, thus meeting the launch vehicle program angle fitting accuracy requirements.

[0087] This invention is not limited to the embodiments described above. Those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications are also considered within the scope of protection of this invention. Contents not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A launch vehicle program angle fit endpoint optimization method, comprising: The application relates to a method for fitting a standard trajectory, and belongs to the field of trajectory fitting. The method comprises the following steps: The second-order difference of the sequence of the program angle with respect to time in the standard trajectory is calculated by using a numerical method, so that the second-order difference of the program angle with respect to time at each time in the input time sequence is obtained; Num time points with relatively large absolute values of the second-order difference are screened out; the time points with relatively large absolute values of the second-order difference are combined, the fitting precision corresponding to each combination is calculated, and the combination with the highest fitting precision is obtained; The maximum angle deviation corresponding to the combination with the highest fitting precision is calculated; if the angle deviation does not meet the precision requirement, the number of fitting end points is increased to recombine and optimize; 2. The launch vehicle program corner fit endpoint optimization method of claim 1, wherein, If the combination with the highest fitting precision meets the angle deviation precision requirement, the jump value of each segment fitting end point is calculated; if the jump value of a certain end point does not meet the precision requirement, the end point is adjusted until the jump values of all end points meet the precision requirement, and the final end point is obtained. Let the sequence of program angles be The corresponding time is [t1t2...t n ], and the first-order difference is The second-order difference is The corresponding time after the second-order difference is [t2 t3... t n-1 ].

3. The launch vehicle program corner fit endpoint optimization method of claim 2, wherein, The second-order difference of the sequence of the program angle with respect to time in the standard trajectory is calculated by using a numerical method, so that the second-order difference of the program angle with respect to time at each time in the input time sequence is obtained; The time points with relatively large absolute values of the second-order difference are screened out; the time points with relatively large absolute values of the second-order difference are combined, the fitting precision corresponding to each combination is calculated, and the combination with the highest fitting precision is obtained; S21, an initial value is assigned to the number of points Num, and the number of screened points Num_out is set; S22, the absolute value of the array ddPhi is obtained to obtain the first Num values in the array |ddPhi|, which are recorded as the array Arr_ddPhi, and the time points corresponding to the Num values are recorded as the array Arr_TddPhi; S23, all the numbers in the array Arr_ddPhi are traversed, each number is compared with the numbers within T seconds, and the number with a smaller value in each comparison is recorded; S24, after the comparison, all the numbers with smaller values recorded are removed to obtain all the elements in the array Arr_ddPhi with larger values and without repeated points within T seconds; 4. The launch vehicle program corner fit endpoint optimization method of claim 3, wherein, S25, if the number of elements screened out by the step S24 is less than Num_out, Num is set as Num+K1, K1 is an integer between 5 and 50, and the step S22 is returned to re-screen until the number of elements screened out by the step S24 is not less than Num_out, the combination of the time points corresponding to the screened elements is recorded as the array Arr_TddPhi_out, and the number of elements is recorded as n1. The time points with relatively large absolute values of the second-order difference are combined, the fitting precision corresponding to each combination is calculated, and the combination with the highest fitting precision is obtained; S27, select m1 elements from Arr_TddPhi_out to combine, get a total of A combination; S28, to Arr_TddPhi_out A combination is traversed, and a fitting precision performance index corresponding to each combination is calculated, including: S281. Assume the current combination is [tp1 tp2 ... tp m1 ], will [tp1 tp2 ... tp m1 [t1 t2 ... t] as endpoint pairs n Divide into segments; S26, an initial value of the number of screened time points m1 is set; S283, obtain [t1 t2... t n ] the fitting program angle result corresponding to each point, record the result as S284, find the fitting performance index under the current combination S29, the fitting performance indicators of all combinations are obtained respectively The minimum value is obtained, and the corresponding combination is the combination with the highest fitting accuracy.

5. The launch vehicle program corner fit endpoint optimization method of claim 4, wherein, S282, the sequence of the program angle Phi is fitted by using a cubic polynomial for each segment; The maximum angle deviation corresponding to the combination with the highest fitting precision is calculated; if the angle deviation does not meet the precision requirement, the number of fitting end points is increased to recombine and optimize; S31, find the maximum value in the combination with the highest fitting accuracy ; S32, if the maximum value in the middle exceeds the allowed range of fitting deviation then let m1 = m1 + K2, where K2 is an integer between 1-5, repeat steps S22-S32 to restart the combination optimization.

6. The launch vehicle program corner fit endpoint optimization method of claim 5, wherein, The combination with the highest fitting accuracy satisfies the angular deviation accuracy requirement, and the jump value at each fitting endpoint of the combination is calculated. If the jump value at a certain endpoint does not satisfy the accuracy requirement, the endpoint is adjusted until the jump values at all endpoints satisfy the accuracy requirement, and the final endpoint is obtained, including: S41, traversing the endpoint values of the combination with the highest accuracy in ascending order of time, and solving each endpoint with the polynomials obtained by fitting the two segments before and after the endpoint, and calculating the absolute value of the difference between the two fitting results, which is the jump value; S42, if the hop value of a certain endpoint currently traversed is greater than the maximum hop value adjusting the endpoint value, including: S421, moving the endpoint value by x steps in the time interval of ±y, and keeping other endpoint values unchanged, to obtain z adjustment items; S422, after adjusting the endpoint, a total of z+1 combinations are obtained, and the fitting performance indicators of the z+1 combinations are calculated; S423, find the minimum value of the performance indicators of the z+1 combinations, and the adjusted endpoint corresponding to the minimum performance indicator is the final endpoint.

7. The launch vehicle program corner fit endpoint optimization method of claim 6, wherein, The x is 0.01-0.1, and the y is 0.1-1.

8. The launch vehicle program corner fit endpoint optimization method of claim 6 or 7, wherein, The fitting deviation allowed range is 0.02-0.1, the maximum jump value is 0.01-0.05.

Images